The 2025 reconciliation law imposes significant new restrictions on states’ ability to generate Medicaid provider tax revenue, including prohibiting all states from establishing new provider taxes or from increasing existing taxes as well as reducing existing provider taxes for states that have adopted the Affordable Care Act (ACA) Medicaid expansion. Medicaid is jointly financed by the federal government and the states, with the federal government guaranteeing states federal matching payments with no pre-set limit. In federal fiscal year (FFY) 2024, the federal government paid 65% and states paid 35% of total Medicaid costs. States are permitted to finance the non-federal share of Medicaid spending through multiple sources, including state general funds, health-care related taxes (referred to as “provider taxes” throughout this brief), and local government funds.

Changes to provider tax rules will have significant effects on state budgets and may make it difficult for states to maintain current Medicaid spending without increasing state general fund spending; but could increase transparency around Medicaid financing. The changes come at a time when states are already experiencing overall slower revenue growth, and it is unclear how states will be able to make up the lost revenues. The changes could exacerbate existing state budget challenges and result in lower provider payment rates or reductions in Medicaid benefits or coverage, although the effects will vary by state.

This issue brief uses data from KFF’s 2025-2026 survey of Medicaid directors and from a proposed rule on provider taxes to describe states’ current provider taxes, explore how rules governing provider taxes are changing because of the 2025 reconciliation law and the regulations implementing that law, and summarizes which changes may affect each state.

1.How have states used provider taxes to help finance the state share of Medicaid?   

KFF’s 2025 Medicaid Budget Survey found that the majority of state Medicaid spending came from general fund revenues, but provider taxes contributed 18%. States have considerable flexibility in determining how to finance the state (or non-federal) share of Medicaid payments, within certain limits. Across all states, most of the state share of Medicaid spending comes from state general funds, but there is considerable variation in how much states rely on other funding sources. KFF’s 2025 Medicaid budget survey found that general funds accounted for a median of 70% of the non-federal share in state fiscal year (FY) 2026 enacted budgets, while provider taxes accounted for 18%, and funds from local governments or other sources accounted for 6% (this is relatively similar to 2018 data on non-federal share funding sources reported by the Government Accountability Office (GAO) and 2024 data on general fund spending from the National Association of State Budget Officers (NASBO)). 

All states but Alaska finance part of the state share of Medicaid funding through at least one provider tax and 41 states have three or more provider taxes in place (Figure 1). Medicaid provider taxes are defined as those for which at least 85% of the tax burden falls on health care items or services or entities that provide or pay for health care items or services (see Social Security Act, Section 1903(w)(3)(A)). Provider taxes may be imposed as a percentage of provider revenues or using an alternative formula such as a flat tax on the number of facility beds or inpatient days. States use provider tax revenues to fund Medicaid “base” rates and supplemental payments; to finance eligibility expansions, including the ACA Medicaid expansion; or to more generally support the Medicaid program. Over time, states have increased their reliance on provider taxes, with expansions often driven by economic downturns or a desire to fund eligibility expansions or provider reimbursement increases. Beyond helping finance the state share of Medicaid, permissible tax arrangements may have potential financial benefits for providers who are subject to the tax and serve a high volume of Medicaid patients.

Provider taxes are most common for institutional providers. That includes hospitals (47 states), nursing facilities (45 states), and intermediate care facilities for people with intellectual or developmental disabilities (33 states, Figure 2). Provider tax revenues often finance supplemental payments to institutional providers, which may be a major source of revenues for those providers. Payment policies vary considerably by state, and research has shown that Medicaid base payment rates are below those of Medicare and often below hospitals or nursing facilities’ costs of providing services to Medicaid enrollees, causing some states to rely more heavily on supplemental payments than others to help cover costs. Beyond institutional providers, states have taxes on managed care organizations (MCOs) (22 states), ambulance providers (21 states), and “other” provider types (9 states) such as ambulatory care facilities and home care providers. Provider tax revenues are most likely to be near the 6% safe harbor limit (described in more detail below) for nursing facilities followed by hospitals and intermediate care facilities for people with intellectual or developmental disabilities (Figure 2).

CMS estimates that states will collect nearly $100 billion in provider tax revenues in 2026, mostly from hospital taxes (Figure 3). Historically, the federal government did not provide consistent or comprehensive publicly available data about states’ provider tax policies or revenue collections. In June 2024, the Medicaid and CHIP Payment and Access Commission (MACPAC), called for increased transparency over how states financed the non-federal share of Medicaid payments. However, CMS requested additional data from states about their provider taxes in 2025 and 2026. Using those data and their own projections, CMS estimates that tax revenues in calendar year 2026 will be $98.6 billion, with $61.8 billion coming from taxes on hospitals and $28.1 billion coming from taxes on managed care organizations.

2. What federal rules governed provider taxes before the 2025 reconciliation law?

Since the 1990s, federal rules governing provider taxes have included three core components—requiring taxes to be “broad-based,” “uniform,” and not hold providers “harmless.” Provider taxes were established in the 1980s, but particularly aggressive use of provider taxes following their establishment in the 1980s led to statutory and regulatory limitations beginning in the 1990s. Federal rules prior to passage of the 2025 reconciliation law specified that provider taxes must be:

• Broad-based, which means the tax is imposed on all providers within a specified class of providers (e.g., the tax cannot be imposed only on providers that see primarily Medicaid patients);
• Uniform, which means the tax must apply equally to all providers within the specified class (e.g., the tax rate cannot be higher on Medicaid revenue than non-Medicaid revenue); and
• Not hold taxpayers (providers) “harmless,” which means states are prohibited from directly or indirectly guaranteeing that providers will receive their tax costs back (i.e., be “held harmless”).

To ensure tax programs are “broad-based,” CMS has specified 19 classes of providers (see 42 CFR Section 433.56). States may obtain “uniformity waivers” of the requirements that taxes be broad-based and uniform if the state can prove the net effect of the tax is “generally redistributive,” and the amount of tax is not directly related to Medicaid payments. In assessing whether provider taxes comply with federal laws, regulations specify that the hold harmless requirement does not apply when the tax revenues comprise 6% or less of net patient revenues from treating patients (see 42 CFR Section 433.68), a level sometimes referred to as a “safe harbor” or “hold harmless” limit.

3. What are new federal rules related to provider taxes?

Changes in the 2025 Reconciliation Law

The 2025 reconciliation law, signed by President Trump on July 4, 2025, imposes significant new restrictions on states’ ability to generate Medicaid provider tax revenue. The Congressional Budget Office (CBO) estimated provider tax policy changes from the 2025 reconciliation law would reduce federal Medicaid spending by $226 billion between 2025 and 2034. Those savings reflect the following changes to federal rules:

• An effective prohibition on new provider taxes or increases to existing ones ($89 billion in savings). The law effectively prevents the enactment of any new provider taxes by establishing a hold harmless limit of 0% for any taxes that were not in effect as of July 4, 2025. It also prevents any increases to existing provider taxes, which are capped at their rates as of July 4, 2025.
• Reduced limits on provider taxes in states that adopted the Affordable Care Act (ACA) Medicaid expansion ($102 billion in savings).  Beginning in FFY 2028, the law gradually reduces the hold harmless limit for states that have adopted the ACA expansion by 0.5% annually until the safe harbor limit reaches 3.5% in FFY 2032. The new limits apply to all provider taxes except for those levied on nursing facilities and intermediate care facilities.
• Revisions to the conditions under which states may receive uniformity waivers ($35 billion in savings). Effective July 5, 2025, the law prohibits states from using uniformity waivers if the tax charges higher or lower rates based on the volume of Medicaid revenues or patients.

Changes in the July 2026 CMS Proposed Rule

CMS released a proposed rule in July 2026 to implement the hold harmless provisions in the 2025 reconciliation law, which included some provisions not required under the law. The list below highlights some of the key provisions included in the proposed rule:

• Broader interpretation of taxes “in effect.” Under the proposed rule, CMS would recognize provider taxes as being in effect as of July 4, 2025 or earlier if they had been enacted. This is less restrictive than guidance issued in November 2025, which also required that all applicable uniformity waivers to have been approved by July 4, 2025 and for states to be “actively collecting” revenues as of July 4, 2025.
• Establishing health insurers as a permissible class of providers. Although not required by the 2025 reconciliation law, the rule proposes to add a new “health insurer” provider class to expand CMS oversight of health-care related taxes that goes beyond the existing “MCO” provider class including Health Maintenance Organizations and Preferred Provider Organizations. The proposed rule does not define the new class, but CMS is seeking comments on the potential scope of the new provider class. Such taxes would be subject to all other requirements governing Medicaid provider taxes, including new limits in the 2025 reconciliation law. CMS notes that these taxes are often imposed through state insurance commissions or departments.
• Discontinuation of the 75/75 test. The 2025 reconciliation law did not address the “75/75” test, under which taxes exceeding the hold harmless limit could remain permissible as long as more than 75% of taxpaying providers do not receive more than 75% of the cost of the tax back through enhanced Medicaid or other state payments. Starting in FFY 2027, the proposed rule would discontinue the 75/75 test.
• Enhanced Reporting Requirements and Compliance System. The proposed rule would significantly expand state reporting requirements and introduce retrospective CMS review to determine ongoing state compliance with the new hold harmless limits. States would be required to submit data (interim in 2026 and final in 2028) to CMS to determine applicable hold harmless limits as of July 4, 2025. States would also be required to submit quarterly reports beginning October 1, 2026, supplying tax collection data as well as information on how tax revenues are used by the state and whether public providers are exempt from the tax.

Under the proposed rule, CMS estimates that federal Medicaid spending would decrease by $246 billion over the next ten years (2026- 2035). This is similar to the CBO estimates although the CBO estimate does not include effects for the year 2035, which accounted for $44 billion of CMS’ total federal spending reduction. While the difference in estimates is relatively small, there are a few key differences in the agencies’ assumptions:

• States response. CBO assumed that states would replace 50% of the lost provider tax revenues with other funding sources but CMS assumes they will only replace 30%.
• Coverage loss. CBO estimated that provider tax changes in the 2025 reconciliation law will increase the number of uninsured people by 1.2 million by 2034, but CMS estimates that there will be no enrollment loss associated with the loss of revenues. 

4. Which states may face reductions in existing provider tax revenues?

States that have adopted the ACA Medicaid expansion and have certain provider taxes above the new hold harmless limits will face reductions in existing provider tax revenues. KFF data show that an estimated 31 states will have to reduce one or more provider taxes on hospitals, MCOs, or ambulances because of the lower hold harmless limits in ACA expansion states (Figure 4). Additional states are likely to be affected because of taxes on other classes of providers. Hospital taxes are the most frequently affected, with 28 of the 31 affected states having a hospital tax over 3.5% of net patient revenues as of July 1, 2025. Over half of the Medicaid provisions in the 2025 reconciliation law apply only to ACA expansion states, including the lower hold harmless limits. Those changes—coupled with lower provider tax revenues—may make it particularly difficult for ACA expansion states to navigate a challenging fiscal climate and increasing numbers of uninsured residents.

If CMS’ proposed regulation is finalized with the new health insurer provider class, additional states will be affected, though it is unclear how many states currently have such taxes in place. CMS’ decision to establish health insurers as a provider class for the purposes of Medicaid provider tax rules means that additional taxes will be subject to new hold harmless limits and in ACA expansion states, additional taxes may be subject to the decreasing hold harmless limits over time.

5. Which states may need to rework their “uniformity waivers?”

Uniformity waivers have allowed states to waive the requirement that provider taxes be broad-based and uniform if CMS determines that the tax is “generally redistributive.”  Provider taxes established through such waivers have generally taxed some types of providers within a class more heavily than others. States may use uniformity waivers to achieve policy goals such as limiting tax burdens for sole community hospitals, rural hospitals, or other vulnerable providers; but states have also used the waivers to impose taxes primarily on Medicaid providers. The disproportionate taxation of Medicaid providers has raised CMS concerns, including during the Biden Administration, and in May 2025, the Trump Administration released a proposed rule that aimed to address those concerns. The final rule was published on February 2, 2026 (Box 2).

The 2025 reconciliation law prohibits states from using uniformity waivers if the tax charges higher or lower rates based on the volume of Medicaid revenues or patients. The law specifies that taxes may not be considered generally redistributive if the state effectively varies tax rates based on the providers’ Medicaid revenues or patients, even if the tax does not explicitly name “Medicaid” when establishing the tax rates. The requirement is largely targeted at MCO taxes but may also apply to other provider tax types. It is effective as of July 5, 2025, but the Secretary may give states up to three fiscal years to come into compliance. The final rule on uniformity waivers provides states with transition periods that depend on what type of tax the waiver applies to and the most recent date of CMS approval for the waiver. Specifically:

• For taxes on MCOs with a waiver approval within 2 years of April 3, 2026, states have until the end of the current calendar to transition their taxes (this is expected to be the case in California and at least three other states).
• For all other taxes on MCOs, states have until the end of FY 2027 (which in most states, means they would need to be complying by July 1, 2027).
• For taxes on entities other than MCOs, states have through the end of FY 2028 to come into compliance.

States may come into compliance by either submitting a new waiver proposal that meets the new requirements from the final rule (Box 2) or they may otherwise modify their tax such that no waiver is necessary.

The final rule states that new limits on uniformity waivers will affect at least nine taxes in at least seven states, with effects starting as early as January 1, 2027 (Figure 5). CMS did not identify the specific states in the final rule, but in the proposed rule, CMS specifically named California, Massachusetts, Michigan, and New York as being affected.  KFF and other researchers expect that the other three states are Illinois, Ohio, and West Virginia. The final rule states that existing MCO taxes would now be prohibited in seven states unless the taxes were modified, and that within those seven states, there were at least two additional taxes affected, including one on hospitals and one on nursing homes. However, elsewhere, in the preamble to the final rule, CMS indicated that there were two nursing facility taxes that would now be prohibited. (It’s unclear whether the second nursing facility tax is within the seven states or in an eighth state.) CMS indicates that additional taxes may need to be modified or eliminated, but it is unknown which states have such taxes or what types of providers the taxes pertain to. Beyond uncertainty surrounding the scope of affected taxes, much remains unknown about how states may respond to the new rule.

The use of prior authorization by health insurers has come under increased scrutiny in recent years, as nearly 7 in 10 insured adults say prior authorization is a burden when it comes to getting health care. Health insurers use prior authorization to reduce the use of low-value or unnecessary care, and in the complex and fragmented U.S. health care system is one of the primary tools used to restrain costs. Nevertheless, this practice may result in delays or denials in receiving necessary care, as well as administrative burdens for patients and providers.

In 2024, the Centers for Medicare and Medicaid Services (CMS) published a final rule (“2024 regulation”) aimed at streamlining and automating the prior authorization process for almost all insurance programs it oversees: Medicare Advantage, Medicaid and CHIP fee-for-service and managed care plans, and Affordable Care Act (ACA) Marketplaces on the federally facilitated exchanges (HealthCare.gov). As part of these requirements, every year, payers in these programs are now required to publicly post on their websites specified prior authorization metrics, including approval and denial rates and response times, aggregated for all medical items and services (excluding prescription drugs) for the previous calendar year.

Insurers were required to post the first year of data, for the 2025 calendar year, by March 31, 2026. KFF collected and analyzed these metrics from the largest insurers in most market segments for which reporting is required, including Medicare Advantage, Medicaid managed care, and the Affordable Care Act (ACA) federally facilitated Marketplace. This issue brief summarizes the data by market segment and insurer and discusses the challenges interpreting the data as well as emerging policy changes that may address some of them.

• Medicare Advantage insurers denied 12% of standard prior authorization requests, Medicaid managed care insurers denied 14%, and ACA Marketplace insurers denied 18%. Denial rates for expedited requests were slightly lower. However, the aggregated approval statistics do not allow for further analysis into which services are being denied.
• Prior authorization denials are rarely appealed, but when they are, a considerable share are overturned. Sixty-seven percent of prior authorization denials were overturned upon appeal in Medicare Advantage, 47% were overturned upon appeal in Medicaid managed care, and 43% were overturned in the ACA federally facilitated Marketplace.
• Median response times in all markets were about 1 day for standard prior authorization requests, and about half a day for expedited prior authorization requests for Medicare Advantage, and approximately 1 day for Medicaid managed care and the ACA Marketplace. Insurers are not required to report response time ranges or differences by service category.  
• Despite the regulatory intent to make insurer prior authorization practices more transparent, difficulty locating and interpreting metrics on insurer websites and gaps in how (e.g., a standardized template that insurers are required to use) and what metrics (e.g., denominators and breakouts by service category) must be reported limit the usability of this information directly by the public.

The regulation states that the overall goal of this public reporting requirement is to promote “accountability for payer prior authorization practices” and envisions consumers using the data when shopping for health insurance. The regulation requires payers to which it applies — Medicare Advantage organizations, Medicaid and CHIP fee-for-service and managed care plans, and insurers on the ACA federally facilitated Marketplace (FFM), hereafter collectively referred to as insurers — to provide a list of all medical items and services that require prior authorization (excluding prescription drugs) and report the following metrics aggregated for all medical items and services (excluding prescription drugs) for each line of business:

• Percentage of standard and expedited prior authorization requests that were approved and denied
• Percentage of standard prior authorization requests that were approved after appeal (i.e., overturned denials)
• Percentage of all prior authorization requests where review timeframes were extended and the request was approved
• Average (i.e., mean) and median timeframes between submission of a prior authorization request and determination for standard and expedited prior authorization requests

Additional metrics were made optional, including:

• Percentage of expedited prior authorization requests that were approved after appeal
• Percentage of prior authorization requests where decision timeframes were extended and then the requests were approved, broken out by standard and expedited requests
• Percentage of prior authorization requests approved within 7 days for standard requests and 72 hours for expedited requests
• Numeric counts for metrics, including total number of requests and appeals received

The 2024 regulation introduced new prior authorization reporting requirements for some business segments and expanded them for others.

• For Medicare Advantage, insurers have been required to report information about the use of prior authorization – including the number of requests, denials, and appeals – for several years, with data available to researchers for 2019 through 2024. Data on the share of prior authorizations approved, denied, and approved upon appeal are largely duplicative of the information Medicare Advantage insurers already report. The regulation adds new requirements to provide information on response times and to report metrics separately for standard and expedited requests.
• For Medicaid and the ACA FFM, uniform federal reporting requirements for prior authorization are new,² though an existing patchwork of state laws and regulations has required some insurers to report certain prior authorization data at the state level.

Data collected for this analysis represent the largest insurers in each business segment — those with at least 2.5% market share in each of their respective markets: Medicare Advantage, Medicaid managed care, and the ACA FFM.³,⁴ In total, this analysis includes data from 14 unique insurers, representing 25 million Medicare Advantage enrollees (69% of all enrollment), more than 35 million Medicaid managed care organization (MCO) enrollees (54% of all enrollment), and nearly 11 million enrollees in the ACA FFM (hereafter referred to as the ACA Marketplace for simplicity), which included 28 states in 2025 (74% of all enrollment in FFM states). KFF weighted the data by enrollment within each market segment. See the Methodology section for more details.

This analysis includes the share of standard and expedited prior authorization requests that were approved and denied, the share of standard prior authorization requests that were approved after appeal, and the median response times for standard and expedited prior authorization requests. We do not report the share of prior authorization requests approved after extension because some insurers reported the data only for standard requests, while others reported data separately for standard and expedited. Optional metrics are not included because they were not consistently reported by insurers.

Approval and Denial Rates

Across market segments, insurers denied between 12% and 18% of standard prior authorization requests in 2025. On average, 12% of standard prior authorization requests and 10% of expedited prior authorization requests were denied in Medicare Advantage; 14% of standard requests and 12% of expedited requests were denied in Medicaid managed care; and 18% of standard prior authorization requests and 16% of expedited prior authorization requests were denied in the ACA Marketplace (Figure 1). Across all markets, the share of standard prior authorization requests that were denied was similar to or slightly higher than the share of expedited prior authorization requests that were denied.

Prior authorization denial rates varied widely across the largest insurers (Figure 2).

Medicare Advantage: Among the six Medicare Advantage insurers examined, the share of standard prior authorization requests that were denied ranged from 5% (Elevance) to 17% (UnitedHealth Group), and the share of expedited requests that were denied ranged from 3% (Elevance) to 13% (Centene). UnitedHealth Group, Centene, Kaiser Permanente, and Elevance all had somewhat higher denial rates for standard requests than expedited requests, while CVS and Humana had higher denial rates for expedited requests than standard requests. These are similar to the overall denial rates found in a previous KFF analysis across Medicare Advantage insurers.⁵

Medicaid Managed Care: Among the eight Medicaid managed care insurers examined, the share of standardprior authorization requests that were denied ranged from 2% (L.A. Care Health Plan) to 23% (Independence Health Group). For expedited requests, denial rates ranged from 4% (L.A. Care Health Plan) to 21% (CareSource). Three insurers (Centene, Independence Health Group, and UnitedHealth Group) reported higher denial rates for standard requests compared to expedited requests, while three insurers (CVS Health, CareSource, and L.A. Care Health Plan) reported slightly higher denial rates for expedited requests.⁶

ACA Marketplace: Among the eight ACA Marketplace insurers included in this analysis, the share of standard prior authorization requests that were denied ranged from an average of 3% (GuideWell) to 25% (Centene). Denial rates for expedited requests ranged from 3% (GuideWell) to 23% (Centene). Four insurers (Centene, UnitedHealth Group, Blue Cross and Blue Shield of North Carolina, and Molina) reported higher denial rates for standard requests compared to expedited requests, while two insurers (Elevance and Health Care Service Corporation) reported higher denial rates for expedited requests.⁷ The remaining two insurers (Oscar and GuideWell) in this analysis had virtually the same rates of denials for both standard and expedited requests.

Market segments inherently have many differences, and the share of prior authorization requests denied also varied for the same insurer participating in multiple markets. For example, UnitedHealth denied 17% of standard requests in Medicare Advantage, 11% in Medicaid managed care, and 21% in the ACA Marketplace.

Appeals Data

Insurers overturned a considerable share of denials upon appeal, with variation by market. Although enrollees or their providers can appeal denied prior authorization requests, research shows that denials are rarely appealed in Medicare Advantage, Medicaid MCOs, or the ACA Marketplace. However, the prior authorization data demonstrate that when denials are appealed, a considerable share are overturned.

The share of standard prior authorization requests initially denied that were then overturned upon appeal⁸ (referred to in the CMS template as “Requests Approved After Appeal”) was 67% for Medicare Advantage, 47% for Medicaid managed care, and 43% in the ACA Marketplace (Figure 3). We do not present appeals data for expedited requests because that metric is optional, and most insurers did not provide this information. High overturn rates could raise questions about whether the initial request should have been approved or could indicate that the initial request was missing the required documentation to justify the service. Either way, the appeals process can be complicated and time-consuming for providers and patients.

In Medicare Advantage, if the plan upholds its original denial, the case is automatically sent to an independent review entity. In contrast, there is no automatic external medical review of upheld denials in Medicaid managed care and the ACA Marketplace. HHS OIG has suggested that the presence of automatic, independent review could incentivize Medicare Advantage plans to closely review their denials at the first level of appeal and may help explain higher appeal overturn rates in Medicare Advantage compared to Medicaid managed care and the ACA Marketplace.

Appealed prior authorization denials were overturned more often by some large insurers than others (Figure 4).

Medicare Advantage: Across the six largest Medicare Advantage insurers, the share of denied standard prior authorization requests that were approved varied substantially. Denials were overturned upon appeal more than 90% of the time by Centene, while Kaiser Permanente overturned denials upon appeal 40% of the time.

Medicaid Managed Care: Among the Medicaid managed care insurers examined, UnitedHealth Group overturned 81% of denials upon appeal. The remaining insurers overturned between 22% (CVS) and 48% (Molina) of denials upon appeal. (One Medicaid managed care insurer included in this analysis (L.A. Care Health Plan) did not report overturned prior authorization denials as a percentage of standard prior authorization requests that were appealed; therefore, its appeals data are not included here.)

ACA Marketplace: On average, Health Care Service Corporation overturned just 16% of denials upon appeal, while Centene overturned just over half (54%) of denials upon appeal. Six of the eight large insurers reported overturning denials upon appeal less than half the time.

Median Response Times

Median response times for prior authorization determinations were about 1 day for standard requests, with slightly more variation for expedited requests (Figure 5).

The median time between the submission of a prior authorization request and the determination by the insurer, or response time, for standard requests was about 1 day for Medicare Advantage, Medicaid managed care, and the ACA Marketplace (0.9 days or just under 22 hours), substantially less than the federally required maximum time permitted for standard requests.⁹ The median response time for expedited requests was about half a day (0.4 days or just under 10 hours) for Medicare Advantage, just under 1 day (0.8 days or approximately 19 hours) for Medicaid managed care, and 1 day for the ACA Marketplace. Federal regulations in 2025 required insurers to send a determination for standard requests within 14 calendar days for Medicare Advantage and Medicaid managed care and within 15 days for the ACA Marketplace. Determinations for expedited requests had to be sent within 72 hours for insurers in all three markets.¹⁰

Median response times likely reflect (at least in part) the increasing use of technology in the prior authorization process, such as artificial intelligence and electronic prior authorization systems. Still, nearly one-quarter (24%) of insured adults reported in a July 2025 KFF poll that their health insurance company had delayed their ability to get a requested health care service, treatment, or medication in the past two years.And other research and media reports demonstrate that some patients experience much longer response times, sometimes with serious health consequences. In addition to the 2024 regulation’s provisions aimed at streamlining and automating the prior authorization process, in June 2025, a group of health insurers voluntarily pledged to expand real-time responses for prior authorization approvals by 2027.

Median response times for prior authorization determinations varied somewhat among large insurers (Figure 6).

Medicare Advantage: The median response time for standard requests varied from less than 1 day (CVS, Humana, and Kaiser Permanente) to 2 days (Centene). Across the six insurers examined, the response times for expedited requests were similar or slightly shorter than the response times for standard requests.

Medicaid Managed Care: The median response time for standard prior authorization requests varied from 1 day (Centene, Elevance, and Molina) to 4 days (CareSource). Elevance had the same median response time for standard and expedited requests, while the other insurers reported shorter response times for expedited requests.

ACA Marketplace: Among the eight large insurers included in this analysis, median response times for standard requests ranged from less than 1 day (UnitedHealth Group, GuideWell, and Health Care Service Corporation) to 4 days (Molina). All but one insurer reported similar or shorter median response times for expedited requests than standard requests.

Uniform federal reporting requirements represent a step toward better understanding insurer prior authorization metrics across insurance markets. While the data provide new insights, particularly for Medicaid managed care and the ACA Marketplace insurers, gaps remain. Although a target audience of prior authorization reporting is consumers, those who do not have a deep understanding of health insurance terminology and concepts may struggle to interpret these reports. In previous KFF research, one-quarter of insured adults reported difficulty understanding specific terms their health insurer uses. Difficulty locating metrics on insurer websites, an absence of more detailed data, and unclear reporting standards could additionally pose challenges for consumers wishing to use the data directly to compare health insurers, as envisioned in the 2024 regulation. However, intermediaries may aggregate and explain the data to make it more accessible for consumers.

Insurers are not required to report the number of prior authorization requests for each metric (only percentages), which limits the ability to make useful comparisons and gauge the scale of requests, denials, and approvals behind reported percentages. For instance, a low denial rate could translate into a large number of denials if the insurer received a substantial number of prior authorization requests. Conversely, a high denial rate might be seen with a relatively low number of prior authorization requests if many of those requests were denied (which could indicate the prior authorization process is better tailored, rather than a concerning number of denials). In a previous KFF analysis of the use of prior authorization in Medicare Advantage, for example, the number of prior authorization determinations per enrollee was inversely related to the share of requests that were denied (i.e., insurers with more prior authorization requests denied a smaller share of the requests than those with fewer requests). As a result, the absolute number of denied prior authorization requests per enrollee was similar across insurers with different denial rates.

Numeric counts would provide more context for understanding whether an apparently high or low denial rate reflects a meaningful volume of prior authorization requests. Additionally, they are also necessary to distinguish a true zero from missing data or inapplicable statistics. For example, the “Approved upon appeal” metric does not include the number of appeals an insurer received. Insurers that report 0% of appeals were approved are not required to indicate whether they received no appeals or that none of the appeals were approved. Furthermore, insurers were not required to report the appeal outcome by service or the reason(s) for the decision, making it difficult to interpret differences across and within markets.

Making comparisons between insurers also requires an understanding of policy nuances.  An insurer might have a high prior authorization approval rate because there are many services requiring prior authorization that are almost always approved. Some insurers exempt providers from at least some prior authorization requirements if they consistently have a high prior authorization approval rate (e.g., 90%), known as “gold card” programs. These programs can further complicate comparisons, as the approval rate no longer reflects the entirety of the procedures that might otherwise have required prior authorization. Insurers that rely heavily on gold carding may, in turn, report lower approval rates: since providers that are most clinically compliant are exempt from prior authorization, reported statistics from these insurers include only providers that may inherently have lower approval rates.

Comparisons between issuers are also challenging because of inconsistencies in how metrics are reported. Although CMS published an example template for reporting the prior authorization metrics, its use by insurers is not mandatory, and most insurers used different formats, which could make comparing metrics across insurers burdensome for the public. Some inconsistencies in reporting level could make comparisons difficult. For example, some ACA Marketplace insurers reported prior authorization metrics at the national level, as opposed to separating their reporting by state, so it is not possible to understand geographic variation in prior authorization practices from the same insurer. A national average denial rate for one insurer may or may not reflect the insurer’s denial rate in the state where the consumer lives.

Response times are difficult to interpret. Comparing response timeframes by type of request can provide some insight into the extent to which insurers are meeting the regulatory timeframe requirements. However, insurers are only required to report median and average response times, which limits the ability to determine how often decisions exceed those timeframes. Also, without a required reporting unit, some insurers reported time in hours, while others used days or fractional days. Although median response times across all three markets were typically well below the required maximums, additional information — such as optional metrics on the percentage of requests approved or denied within the required timeframe and ranges of response times — would allow for a more complete picture of how quickly prior authorization decisions are made and how frequently decision times exceed the required timeframes. Additionally, consumers may not always understand how to interpret the median – that is, by definition, half of the prior authorization requests exceed the median time reported.

Comparing insurers across lines of business requires consideration of differences between enrollee populations. Case mix (how sick the enrollee population is) and service mix (which services are most commonly provided for enrollees) likely vary substantially between lines of business, and potentially between insurers. For example, the older Medicare Advantage population likely differs substantially from the general population covered by ACA Marketplaces and the higher-need populations of Medicaid. If denial rates vary by service type and the services for which denials are more common comprise a larger share of prior authorization requests in a given business segment, the higher denial rate would reflect that difference in service use. However, without more detailed information, it is not possible to identify what is driving the differences.

In July 2026, CMS published an updated metrics reporting overview and template for insurers with changes that address some of the gaps addressed above for the 2027 reporting period. It specifies that posting metrics in locations on insurers’ websites that cannot be reached through ordinary navigation is not considered to meet the regulation’s requirement that the information be “publicly accessible.” While still not required, the updated overview does stress that the numerators and denominators specified in the template are “expected.” It notes that median determination response times that are less than one day must be reported in hours, not rounded to “0 days.” It also recommends that insurers identify and explain in their report any data quality issues or concerns.

In addition, the 2026 CMS Interoperability Standards and Prior Authorization for Drugs proposed rule (“2026 proposed rule”) attempts to address some of the data gaps in the 2024 regulation by expanding the prior authorization data that must be publicly reported. The proposed rule would require the same CMS-regulated insurers to publish new metrics for non-drug items and services (e.g., for both standard and expedited prior authorization requests, the number and percentage of requests that remain denied after appeal), as well as the numeric counts for both the new and existing metrics. Also, while CMS currently provides suggested denominators (or universe of prior authorization determinations to include), the proposed rule would standardize the denominators that insurers must use when calculating different metrics. Together, these changes would provide more insight into prior authorization requests, denials, and appeals; make the data easier to interpret; and improve the ability to make comparisons across insurers. More detailed prior authorization metrics could help assess whether initiatives to improve the prior authorization process more broadly, such as the voluntary commitments made by several insurers in June 2025, are leading to meaningful change. However, the proposed rule does not include a requirement that insurers report prior authorization metrics by service category.

Another significant proposed change would require insurers to report prior authorization metrics for prescription drugs, separate from non-drug items and services, for both standard and expedited requests. (This requirement does not apply to insurers offering Medicare Part D prescription drug plans, where separate reporting requirements for coverage determinations already apply.)

There are additional emerging federal and state data reporting policies that are market-specific:

Medicare Advantage

CMS has announced a pilot program where Medicare Advantage insurers can voluntarily submit more detailed prior authorization data, including data for specific service categories. The Trump administration has indicated that it intends to make more detailed reporting mandatory starting with the 2027 plan year. While that may help understand the impact of prior authorization in Medicare Advantage, the data will not be available for several years.

Federal legislation entitled “Improving Seniors’ Timely Access to Care Act of 2025” aims to improve prior authorization in Medicare Advantage. The bill would require Medicare Advantage plans to report certain prior authorization metrics annually to the Secretary of Health and Human Services, who would publish them on a public website. Some required metrics are similar to those required by CMS under the 2024 regulation, but the bill would also require plans to report the percentage and number of specified requests denied or approved during the previous plan year through the use of technology (e.g., artificial intelligence technology).

Medicaid Managed Care

Beginning July 2026, CMS requires states to collect and report prior authorization metrics for Medicaid managed care plans in Managed Care Program Annual Reports (MCPARs) submitted to CMS. In addition to the metrics required by the 2024 regulation, states are required to include the total number of standard and expedited prior authorization requests received. CMS makes MCPARs publicly available on Medicaid.gov, which may make data easier to locate in the future, though there will be a lag between state submission and CMS public posting.

The 2026 proposed rule would align prior authorization metric reporting for Medicaid and CHIP managed care with contract rating periods (which vary across states) and would require Medicaid and CHIP managed care plans to publicly post metrics no later than 90 days after each rating period ends. The rule would also require plans to report metrics by program (defined by a specified set of benefits and eligibility criteria identified in managed care plan contracts), as well as by plan. This proposal aligns with MCPAR reporting requirements and could provide more meaningful information on prior authorization performance by plan and program than aggregate information across programs within a state.

Private Insurance Market

Federal efforts to increase transparency into prior authorization practices in commercial coverage have so far been limited to ACA plans on the FFM. There is no publicly available prior authorization data for self-funded plans sponsored by private employers, which cover the majority of workers with employer-sponsored health insurance. Some states have taken action to increase transparency into prior authorization practices, including by collecting more granular prior authorization data from state-regulated insurers than the 2024 federal regulation requires and using the data to inform future actions aimed at curbing certain insurer practices.

For example, in 2025, Massachusetts conducted a special examination requiring insurers offering fully insured health plans in the state’s private insurance market to submit detailed prior authorization data (including numeric counts and prescription drugs) broken out by service category. The state published an analysis of the data in 2026, noting that it intended to use the data to inform future regulatory guidance. Following this examination, the state updated its insurance regulations to prohibit prior authorization requirements for a wide variety of services provided in-network, including some identified in the state’s analysis as having the most prior authorization requests and high approval rates, such as radiology (for enrollees with cancer) and physical, occupational, and speech therapy.

Similarly, in 2025, Iowa enacted a law requiring state-regulated insurers to submit certain prior authorization data to the state annually, which will be made available to the public, and requiring utilization review organizations to eliminate prior authorization requirements for health care services that meet certain criteria.

Going forward, more states may build upon the federal requirements by requiring insurers in their state to provide more detailed information about prior authorization practices and use that information to address the scope of prior authorization requirements.

This analysis includes data collected from the websites of Medicare Advantage, Medicaid managed care, and ACA federally facilitated Marketplace (FFM) insurers that each had at least 2.5% market share of enrollment in their respective market segments. In total, we collected data from 14 unique insurers in these markets, representing approximately 71 million enrollees. See Appendix Table 1 for more details.

We reported prior authorization metrics at the parent company level. Subsidiary insurers may have a different name from the parent company, and the parent company may own more than one insurer. For ease of reading, we refer to the parent company as the “insurer” in this analysis.

Medicare Advantage

Public reports were collected for Medicare Advantage insurers enrolling at least 2.5% of all Medicare Advantage enrollees in 2025. Altogether, these insurers comprised 72% of Medicare Advantage enrollment in 2025. In some cases, reports could not be located, and so the data included in this analysis reflects information provided for Medicare Advantage contracts covering 69% (25 million) of Medicare Advantage enrollees in 2025. Specifically, reports covering at least 95% of enrollment were identified for each insurer included in the analysis, with the exception of Elevance-sponsored plans, where reports covering 83% of enrollment were identified. Medicare Advantage organizations are required to report metrics at the contract level. CMS December 2025 enrollment was used to weight contract-level data when aggregating to the insurer and total market level.

Medicaid Managed Care

Public reports were collected for Medicaid MCO insurers enrolling at least 2.5% of all Medicaid MCO enrollees as of July 2024 (the latest available national Medicaid managed care enrollment data). Reports for limited benefit prepaid ambulatory health plans (PAHPs) and prepaid inpatient health plans (PIHPs) were not included. Altogether, these insurers comprised 57% of Medicaid MCOs as of July 2024. In some cases, reports could not be located, and so the data included in this analysis reflects information provided for Medicaid MCO contracts covering 54% (nearly 36 million) of Medicaid MCO enrollees as of July 2024. Specifically, reports covering at least 92% of enrollment were identified for each insurer included in the analysis, with the exception of Molina and Independence Health Group plans, where reports covering 87% and 84% of enrollment, respectively, were identified. CMS July 2024 enrollment was used to weight plan-level data when aggregating to the insurer and total market level.

Although the 2024 regulations require Medicaid and CHIP fee-for-service (FFS) programs and CHIP managed care plans to publicly report prior authorization metrics, these entities are not included in this analysis because reports were difficult to locate and comprehensive national plan-level enrollment data are not available for CHIP managed care. Additionally, the populations and services covered under FFS vary substantially across states, limiting comparability.

ACA Marketplace

We searched for public reports for the nine ACA Marketplace insurers with at least 2.5% market share in the 28 states that used the FFM (HealthCare.gov) in 2025. (Federal reporting requirements do not extend to states that operate their own Marketplaces, including state-based Marketplaces that use the HealthCare.gov platform). Prior authorization metrics could not be located for one of these insurers (CVS, which had 5% market share). UnitedHealth Group, which reported prior authorization data at the state level, did not include data for its plans sold in Oklahoma (less than 1% of its enrollment). Altogether, the data in this analysis reflect information provided for Marketplace insurers covering 74% (10.8 million) of ACA FFM enrollees in 2025.

Marketplace insurers are required to report metrics at the issuer level. Some insurers broke out their prior authorization metrics by state, while others provided national averages. ACA Marketplace enrollment was obtained from Mark Farrah Associates Health Coverage Portal, including mapping insurers to parent companies. To weight insurer-level data to the total market, enrollment was first filtered to FFM states and then aggregated to parent company.

Due to the large number of smaller insurers in the ACA Marketplace, relative to Medicare Advantage and Medicaid managed care, we also collected and analyzed data for ACA FFM insurers that had at least 1% market share in 2025, compiled in Table 2 of the Appendix. For consistency across market segments, however, only data from insurers with at least 2.5% market share were included in the overall analysis.

1. The turnaround time for standard requests in Medicare Advantage and Medicaid managed care was shortened to 7 calendar days beginning in 2026. Timeframes for the ACA Marketplace remain the same. ↩︎
2. CMS makes data on post-service (not pre-service/prior authorization) claims denials available for plans on the ACA FFM. ↩︎
3. Except CVS (which had 5% market share on the ACA FFM in 2025), as the company’s prior authorization metrics for that market could not be located. ↩︎
4. Due to the large number of smaller insurers in the ACA Marketplace, relative to Medicare Advantage and Medicaid managed care, we collected data for ACA FFM insurers that have at least 1% market share, compiled in Appendix Table 2. For consistency across market segments, only data from insurers with at least 2.5% market share were included in this analysis. ↩︎
5. Centene includes Wellcare, Elevance includes Anthem BCBS and Wellpoint, and CVS includes Aetna. ↩︎
6. Elevance includes Amerigroup, Independence Health Group includes AmeriHealth Caritas, and CVS includes Aetna. ↩︎
7. Centene includes Ambetter Health; GuideWell includes Capital Health Plan (FL), BCBS of FL, Health Options (FL), and Florida Health Care Plan; Health Care Service Corp. includes BCBS plans in Montana, Oklahoma, and Texas; Elevance includes Wellpoint and Anthem BCBS. ↩︎
8. FAQs currently on the CMS website indicate that appeals metrics should aggregate all levels of appeals, including internal and external reviews. Note that this clarification appears to have been added in early 2026. ↩︎
9. Some insurers reported response times in days while others used hours. To standardize, we converted hours to days and report time in days. ↩︎
10. The turnaround time for standard requests in Medicare Advantage and Medicaid managed care was shortened to 7 calendar days beginning in 2026. Timeframes for the ACA Marketplace remain the same. ↩︎

• Among states that use state-appropriated funds to pay for abortion for Medicaid enrollees, variability in abortion payment rates for the same procedures is considerable. Twenty-one states currently use their own funds to cover abortion services for Medicaid enrollees beyond the limits imposed by the federal Hyde Amendment, which bans federal funding of abortions in cases other than pregnancies that result from rape, incest, or are a life threat. In these states, Medicaid medication abortion rates range from $162 (RI) to $869 (WA) (median $597); dilation and curettage (D&C) procedure rates go from $126 (RI) to a high of $1,000 (NY) (median $378); and dilation and evacuation (D&E) procedure rates vary from $126 (RI) to $1,920 (IL) (median $636).  
• Only a few states significantly increased their state-funded reimbursement rates for abortion services for Medicaid enrollees in the past two years. Most states that cover abortions for their enrollees have not changed or only modestly increased their reimbursement rates. Among the states that increased D&C rates, only 5 did so at or above the rate of health care inflation. Similarly, among states that increased D&E reimbursement rates, only 6 increased rates at or above the rate of health care inflation over the past two years. Over half (57%, 12 states) have not increased D&C procedure reimbursement rates, and 52% (11 states) have not increased D&E procedure reimbursement rates. 
• Despite the increased complexity of abortion procedures later in pregnancy, Medicaid reimbursement rates are not substantially higher than rates for earlier-stage procedures in most states. Reimbursement rates for abortions later in pregnancy typically do not reflect the increased complexity and costs associated with later-stage care. The median reimbursement rate for states that fund abortions for Medicaid enrollees was $378 for D&C—typically used earlier in pregnancy—vs. $636 for D&E, which is generally used for abortion procedures performed later in pregnancy. These rates are considerably lower than previously published research on self-pay abortion charges for the same procedures.

Since the Dobbs decision in 2022 overturning Roe v. Wade, 13 states have banned the provision of abortion with few exceptions. In the remaining states where abortion is legal, another barrier to abortion services has been the federal Hyde Amendment, which prohibits the use of federal Medicaid funds for abortion with limited exceptions for pregnancies that endanger the life of the pregnant person, or that result from rape or incest. States may use their own revenues to pay for abortion services for Medicaid enrollees beyond the federal financing restrictions, and 21 currently use their own state funds to pay for abortions for Medicaid enrollees, while 16 states & DC where abortion provision is not banned follow the Hyde Amendment restrictions (Figure 1). Medicaid reimbursement rates are established by the states under broad federal guidelines and have historically been lower than those paid by Medicare and are even lower relative to private insurance rates.

In the case of Medicaid, lower reimbursement rates can lead to abortion providers experiencing financial shortfalls, challenging their long-term sustainability and ability to serve Medicaid enrollees. Lower reimbursement rates have been linked to higher out-of-pocket costs for Medicaid beneficiaries, who are overwhelming low-income, further constraining their access to abortion care. 

To understand the status of Medicaid payment for abortion services four years post-Dobbs, KFF researchers reviewed Medicaid physician fee schedules for medication and procedural abortions in states that do not ban abortion, updating an analysis first published in 2024. This analysis focuses on states where abortion provision is not banned, including those using state funds to pay for abortions for Medicaid enrollees as well those that only cover abortions under Hyde exceptions, and it excludes the 13 states where abortion provision is banned. 

Medicaid services are financed via two pathways: fee-for-service (FFS) and managed care arrangements. The reimbursement rates presented in this brief are for FFS claims, as contracted managed care rates are not typically publicly available. The FFS reimbursement rates reported in this brief are for non-facility (e.g., outpatient clinics or physician offices), provider-only rates (e.g., excluding any facility rates), since most abortions are performed outside of a hospital setting. States such as Connecticut, Massachusetts, Nevada, and Washington use different reimbursement fee schedules for family planning and abortion clinics. These reimbursement rates are often higher than physician fee schedules. This analysis uses the family planning and abortion clinic rates, when available.

State Medicaid programs also differ in the payment structures they use for abortion services. Some states use a bundled reimbursement structure for abortion services, where ancillary services provided alongside the abortion are included in the bundled payment rate. Other states use unbundled billing, and providers can bill separately for all services provided with the abortion. Additional services often billed for on the day of the abortion in states that use unbundled codes may include an ultrasound, administered medication, a nerve block, and Rh testing, which are outlined in coding guides developed by the Reproductive Health Access Project for manual vacuum aspiration abortion and medication abortion. Median reimbursement rates for each of these services from state Medicaid physician fee schedules are reported in the tables throughout this brief.

Even when Medicaid reimburses for abortion services, providers have reported that Medicaid reimbursement rates are substantially lower than what they receive from self-pay patients. Because there are limited and outdated data on the actual cost of providing abortion care, self-pay prices can serve as a useful proxy for provider costs, although they may underestimate true costs if they reflect discounts provided to patients by abortion funds. Research conducted before the Dobbs decision illustrates the breadth of this reimbursement gap. A 2020 study reviewed 2017 Medicaid and Medicare physician fees schedules for D&C and D&E procedures across 45 states and D.C. and found that median Medicaid reimbursement rates for first- and second-trimester abortions covered only 37% and 41% of the amount charged to self-pay patients for the procedures, respectively. The study did not address reimbursement rates for abortions after the second trimester. Providers also reported instances in which they received no reimbursement for abortions that were eligible for Medicaid coverage under the Hyde Amendment. The study's authors concluded that this reimbursement gap, particularly when combined with inconsistent reimbursement practices, may further discourage providers from participating in Medicaid.

More recent data on self-pay prices support these findings. A 2024 report on self-pay charges for abortion services from 2017 to 2023 found median charges of $563 for medication abortion, $650 for a first-trimester procedural abortion (D&C), and $1,000 for a second-trimester abortion (D&E). These quoted self-pay prices are substantially higher than Medicaid reimbursement rates in many states, reinforcing that a gap exists between what providers typically charge and what Medicaid pays. 

In 2023, medication abortion, a two-drug regimen using mifepristone and misoprostol that is FDA-approved to terminate pregnancies up to 10 weeks in the U.S., accounted for 65% of all abortions. This share has increased since the FDA removed the in-person dispensing requirements in 2021, making medication abortion via telehealth, mail, and pharmacies more accessible. 

Medication abortion can be billed using three separate procedure codes, and often all three codes are billed at the same time. There are codes for two pharmaceuticals—mifepristone (HCPCS S0190) and misoprostol (HCPCS S0191)—as well as a global medication abortion code (HCPCS S0199) that includes all affiliated services and supplies (e.g., patient counseling, office visits, confirmation of pregnancy by HCG, ultrasound to confirm duration of pregnancy, ultrasound to confirm completion of abortion). 

States typically take two approaches to reimburse for medication abortion services: (1) a bundled payment using the global medication abortion code plus the medications (billed either through medical services or pharmacy billing channels); or (2) payment for separate services, like office visits and ultrasounds, plus the medications. 

Twenty out of 21 states that use state funds to pay for abortion services for Medicaid enrollees list reimbursement rates for at least one of the three medication abortion codes, with 16 listing rates for the global medication abortion code, 13 listing rates for mifepristone, and 12 listing rates for misoprostol. The one state not accounted for is New York; while they do not reimburse for the global medication code, on their publicly available fee schedule they reimburse at the “cost” of the drugs to abortion providers for both mifepristone and misoprostol. The median Medicaid reimbursement rate for the global medication abortion code is $514, ranging from a low of $81 in Rhode Island to a high of $825 in Washington (Figure 2). The median reimbursement rate for mifepristone is $81, ranging from a low of $43 in Washington to a high of $128 in New Jersey. The median reimbursement rate for misoprostol is $2, which ranges from less than $1 in Washington to $22 in Illinois. When the median amounts for all three medication abortion codes are summed together, the median reimbursement rate for medication abortion is $597. However, the range of summed codes across states is quite large, from $162 in Rhode Island to $869 in Washington.

For the five states that do not use the bundled, global code for medication abortion, the sum of the rates for mifepristone and misoprostol alone range from $44 in Minnesota¹ to $131 in New Jersey. In these states, providers may bill separately for ancillary services, which could potentially add additional reimbursement (Table 1).

The medication abortion reimbursement rates reported in this brief apply to services provided in outpatient clinics or physicians’ offices and do not specifically address reimbursement rates for telehealth medication abortion for either bricks-and-mortar or online-only providers. Telehealth has become an increasingly common mode of medication abortion provision; in 2025, approximately 28% of all abortions were medication abortions provided through telehealth, representing a 27% overall increase from 2024. Although research on Medicaid reimbursement rates for telehealth medication abortion is limited, a 2024 report from the National Health Law Program found that, across six states (Illinois, Minnesota, New Mexico, New York, Rhode Island, and Washington), there was overall payment parity for synchronous video telehealth service delivery under Medicaid compared to in-person medication abortion service delivery. However, many of these states adopted temporary telehealth payment parity policies in response to the COVID-19 public health emergency, and some of those policies have not yet been made permanent, limiting the generalizability of these findings. 

While most states that use their own funds to pay for abortion services for Medicaid enrollees list reimbursement rates for medication abortion, only half of the states that follow Hyde restrictions list reimbursement rates for medication abortion, and median reimbursement in these states is substantially lower (see Appendix Table 1). Beyond the issue of failing to publish reimbursement rates for medication abortion, a 2025 U.S. Government Accountability Office (GAO) report found that 14 states participating in the Medicaid Drug Rebate Program (MDRP) were not complying with federal requirements to cover FDA-approved medication abortion drugs, even in circumstances permitted under the Hyde Amendment, such as rape, incest, or life-endangerment. The GAO reported these instances of noncompliance to the Centers for Medicare and Medicaid Services (CMS) in late 2025. 

Dilation and curettage (D&C) is a common abortion procedure that can be used up to approximately 16 weeks of pregnancy. Medicaid physician fee schedules for fee-for-service reimbursement rates were published online for D&C procedures in all 21 states that use state funds to reimburse for abortion services for Medicaid enrollees beyond Hyde exceptions (Figure 3). The median reimbursement rate for a D&C procedure in these states was $378, ranging widely from $126 in Rhode Island to $1,000 in New York. Since 2024, Pennsylvania² and Colorado have shifted from following Hyde restrictions to using state funds to pay for abortion for Medicaid enrollees. In the past two years, Pennsylvania increased their rate for D&C procedures by 833%, Colorado increased their reimbursement rate by 378%, and Maryland increased their rate by 154%. Beyond these outliers, most states have not increased their rates or increased them only slightly. Nine of 21 states that use state funds to cover abortion for Medicaid enrollees increased rates, but of those, only 5 states increased rates at or above the health care inflation rate over the past two years. Of the remaining 12 states that did not increase rates, 6 states have not changed their D&C reimbursement rates, and 6 states have decreased their rates between 2024 and 2026. In states that only reimburse for abortions in cases of pregnancies resulting from rape, incest, and life endangerment, payment rates were considerably lower (see Appendix Table 2).

Other services may be billed and reimbursed on the day of a procedural abortion including ultrasound, anesthetic medication administration like lidocaine and methergine, or a nerve block for pain. In some states, the reimbursement rates for these additional services are not publicly posted (Table 2). Illinois and New Mexico reimburse procedural abortions (CPT codes 59840 and 59841) as bundled codes and will not pay for other related services when these codes are billed. Therefore, Illinois and New Mexico were removed from the calculations in the table below detailing median ancillary rates for services that may be provided at the time of the abortion. 

For a dilation and evacuation (D&E) procedure, which is often used in the second trimester, the reimbursement rates similarly varied widely by state (Figure 4). In states that fund abortion services for Medicaid enrollees, the median reimbursement rate for a D&E procedure was $636, ranging from a low of $126 in Rhode Island to a high of $1,920 in Illinois. Since 2024, Colorado increased their D&E reimbursement rate by 492%, Pennsylvania increased their rate by 227%, and Massachusetts by 99%. Beyond those substantial reimbursement rate increases, 10 of 21 states that use state funds to pay for abortion for Medicaid enrollees increased D&E reimbursement rates for Medicaid providers between 2024 and 2026, but only 6 increased their rates at or above the rate of health care inflation. Of the remaining 11 states that did not increase rates, 7 have not changed their D&E reimbursement rates, and 4 have decreased their rates over the past two years. As with D&C procedures, the median reimbursement rate for D&E procedures in states that pay for abortions for Medicaid enrollees is higher than in states that limit abortion coverage to those cases permissible under the Hyde Amendment (see Appendix Table 3).

As with D&C procedures, providers will often bill for other services provided with the D&E procedure if the code is not a bundled code (Table 3). In line with D&C procedures, Illinois and New Mexico use bundled rates for D&E procedures that include ancillary services and, therefore, are not included in the calculations in the table below.  

Given that D&E procedures are typically provided later in pregnancy and are more costly and complex procedures, it would follow that reimbursement would be substantially higher than for D&C procedures, which are typically done earlier in pregnancy. The majority of states (27 of 38) without bans—including Hyde and non-Hyde states—reported a Medicaid reimbursement rate difference that was less than $200 (Figure 5). Three states (Rhode Island, Nebraska, & Wisconsin) reimburse for first and second trimester abortions at the exact same rate, failing to account for the differences in cost and complexity of these different procedures, leaving providers to either be exposed to financial losses if they provide abortions later in pregnancy or decline to accept Medicaid patients.

In analyzing the role of Medicaid in abortion financing, the Guttmacher Institute’s 2021-2022 Abortion Patient Survey found 62% of abortion patients living in states that funded abortion beyond Hyde restrictions used Medicaid to pay for their abortion. Comparatively, for abortion patients living in Hyde restricted states, 82% paid out of pocket for their care, with two-thirds of those surveyed reporting they had to raise money to afford their abortion, often through accruing debt, relying on financial subsidies from abortion funds or clinic discounts, or asking friends and family. Abortion funds have long been relied on as a safety-net option for those in need of assistance in paying for their abortion, but a post-Dobbs trend of increased requests for support coupled with decreased donations leaves funds in precarious and depleted financial positions.  

These disparities in how abortion care is financed highlight the fact that insurance coverage alone is insufficient to assure access, particularly when providers’ reimbursement policies inadequately support the costs of the full range of abortion services. While reimbursement for care after the second trimester is not directly addressed in this report, a recent study investigated self-pay prices paid by patients seeking later abortion care and found that prices increased with gestational stage, while, inversely, provider and clinic availability and Medicaid coverage declined. Low Medicaid reimbursement for these services may further exacerbate reproductive access inequities for Medicaid enrollees.  

Beyond implications on patient affordability, reimbursement rates also affect providers’ ability to provide abortions to all patients regardless of payor. As with all health services, the cost of providing abortions has grown in all states as the costs of medical equipment and personnel increase annually. In addition, abortion providers face expenses that are not incurred by most other outpatient clinics such as security costs to keep their staff and patients safe from anti-abortion activities, such as extra security guards, cameras, staff background checks, and bulletproof windows. Increased safety concerns and costs have also made it difficult to retain the abortion clinic workforce.