Medicaid GLP-1 Prior Authorization Criteria by State2026 Data — Dataset Cutoff July 31, 2026

Weight Loss Provider Guide Research is an independent research and reference resource covering access, coverage, and policy for medical weight-management care.

Medicaid GLP-1 prior authorization criteria by state are not converging — they are diverging. Reviewing publicly available state materials source by source, we identified 11 state Medicaid fee-for-service programs with an obesity or chronic-weight-management GLP-1 pathway in force in the materials reviewed through July 31, 2026: Delaware, Kansas, Michigan, Minnesota, Mississippi, Missouri, North Carolina, Rhode Island, Tennessee, Virginia, and Wisconsin. KFF counted 13 in January 2026. Utah's one-year pilot window ended June 30. MassHealth ended coverage of drugs used solely for obesity or overweight effective July 3.

That is the number most people came for. It is also the least interesting thing on this page.

Here is the more useful finding: among the seven active programs for which we could reproduce a current initial authorization period, approval runs from three months in Rhode Island and Tennessee to up to twelve months through Mississippi's electronic DUR+ pathway. The renewal test is not standardized either. Minnesota, North Carolina, and Tennessee use a fixed percentage; Kansas uses a percentage tied to time; Wisconsin requires a documented BMI reduction without publishing a fixed percentage in the criteria reviewed; and Rhode Island measures the first month in pounds. A national count of “11 states” flattens all of it.

July 31 matters for another reason. It was the application deadline for state Medicaid agencies seeking to join the voluntary CMS BALANCE Model. Its state Medicaid terms set standardized access rules for participating states. Those rules are conditional participation terms, not a nationwide mandate, and CMS had not published a participant list by this dataset's cutoff.

Table 1. Medicaid obesity GLP-1 coverage and prior authorization at a glance, by state

Programs with a publicly documented obesity or chronic-weight-management pathway in force in the materials reviewed through July 31, 2026. “Preferred” describes formulary placement, not the absence of prior authorization.

Source: Weight Loss Provider Guide Research, “Medicaid GLP-1 Prior Authorization Criteria by State,” dataset cutoff July 31, 2026. Compiled from current state Medicaid preferred drug lists, prior authorization forms, provider bulletins, manuals, state-plan materials, and enacted policy. Full per-state sources appear below.

“Not publicly stated” means the field was not located in the current statewide materials reviewed. It does not mean no requirement exists. “Not reproducibly established” means the source available at the cutoff did not preserve the exact version needed to support the number.

What this shows, and what it doesn't

It shows dated statewide policy found in public official documents: which programs had a documented obesity pathway in the materials reviewed through July 31, which products were preferred or placed in the obesity class, and the thresholds, durations, and renewal tests that could be reproduced.

It does not show whether any particular person will be approved. Prior authorization is adjudicated member by member against diagnosis, documentation, claim history, benefit channel, and plan-specific edits. A state row describes published policy. It is not a prediction.

It also does not pretend that every managed-care organization inside a state uses the fee-for-service rule. Where a state's own materials indicate plans may differ, the channel stays visible instead of being averaged into a false statewide answer.

Nothing here is medical advice or a coverage determination. For an individual case, the state Medicaid agency or the member's health plan is the coverage authority, and a prescribing clinician is the right person to assess whether a medication is appropriate.

How to read the terms on this page

Source: Weight Loss Provider Guide Research data dictionary, July 31, 2026.

Medicaid GLP-1 prior authorization criteria by state: how many programs cover obesity in 2026?

We identified 11 state Medicaid fee-for-service programs with a publicly documented obesity or chronic-weight-management pathway in force in the materials reviewed through July 31, 2026. KFF counted 13 in January 2026, down from 16 in October 2025. The July result incorporates two later changes: Utah's one-year weight-management pilot window ended June 30, and MassHealth ended coverage of drugs used solely for obesity or overweight effective July 3.

Here is the arithmetic, stated plainly so anyone can check it:

KFF January 2026 fee-for-service baseline               13
  less Utah after the pilot window ended June 30         −1
  less Massachusetts after the July 3 coverage change    −1
Publicly documented active programs at July 31 cutoff    11

The 13-state baseline is KFF's, from its survey of state Medicaid officials. The subtraction and the 11-state result are ours, based on the current Utah and Massachusetts documents.

One honest limitation belongs on the number itself. CMS accepted BALANCE applications through July 31, 2026 and had not published a complete list of accepted or pending state participants by the cutoff. A state decision not yet visible in public program materials would not appear here. That is why the finding says identified in publicly available state materials reviewed instead of claiming that an undisclosed program could not exist.

The 51-jurisdiction status ledger

The national ledger includes all 50 states and the District of Columbia. It contains three different evidence classes and does not pretend they are equivalent.

Source: KFF's January 2026 survey baseline, updated with the state-level changes and active-state review documented on this page. The 34 baseline-only rows are dated survey findings, not individual July verifications.

Why the same national count can still name different states

Two sources can publish the same number and mean different states. Utah is the clean example. Its June 2026 form says weight-loss coverage was part of a pilot that might not continue after June 30 and asks whether the member already had an active PA approved during the July 1, 2025–June 30, 2026 window. We found no new-start obesity-only pathway in the July materials reviewed. Rhode Island points the other direction — its enacted restriction is recorded separately because implementation had not yet been established in a state notice we could reproduce.

The underlying question is ambiguous until three fields are fixed in advance:

1. Indication: obesity only, or any FDA-approved indication.
2. Benefit channel: fee-for-service, managed care, or both.
3. Date: the effective date of the rule, not merely the date a webpage was updated.

Change any one and the count can move.

What the count includes and excludes

We counted a program as active only where a current official source documented an obesity or chronic-weight-management pathway. The national obesity count therefore excludes:

• Type 2 diabetes pathways. Coverage of a diabetes indication does not establish obesity coverage.
• Cardiovascular-risk reduction. Wegovy has a separate FDA-approved cardiovascular indication.
• Moderate-to-severe obstructive sleep apnea. Zepbound has a separate FDA-approved OSA indication.
• Noncirrhotic MASH with F2–F3 fibrosis. Wegovy injection has a separate FDA-approved MASH indication.

Sources: 42 U.S.C. § 1396r-8(d)(2); Wegovy prescribing information; Zepbound prescribing information.

What are the Medicaid GLP-1 prior authorization criteria in each state?

The criteria are not interchangeable. In the current sources we could reproduce, six programs publish a comparable numeric adult pathway, seven publish a current initial authorization period, and four different structures are used to measure continuation. A chart note that satisfies Tennessee may not satisfy Wisconsin, and a product preferred in Minnesota may be nonpreferred in Missouri.

Table 2. Active-state Medicaid GLP-1 criteria matrix, July 31, 2026

Source: Weight Loss Provider Guide Research, dataset cutoff July 31, 2026. Normalized from the official state sources linked in the state notes and primary-source register below. Inequality symbols are preserved as written.

Five things the matrix makes visible that a coverage map cannot

1. 1. The reproducible initial authorization range is fourfold.

   Seven of the 11 active programs publish a current initial period in the sources used here. Rhode Island and Tennessee start at three months. Kansas uses 28 weeks. Minnesota and North Carolina use six months. Wisconsin uses up to 183 days. Mississippi's DUR+ pathway can authorize up to 12 months.

2. "Successful treatment" has at least four structures.

   Minnesota, North Carolina, and Tennessee use a fixed percentage. Kansas uses percentages tied to different time points. Wisconsin requires a documented BMI reduction without a fixed percentage in the criteria reviewed. Rhode Island uses an absolute four-pound first-month measure followed by continued loss.

3. 3. Preferred does not mean authorization-free.

   Delaware's entire obesity class requires PA. Kansas places Wegovy and Zepbound in a preferred column under an anti-obesity PA framework. Rhode Island requires manual PA for preferred agents for recipients 21 and older. Wisconsin marks all drugs in the class as PA-required.

4. 4. A state-level "yes" does not tell you which drug.

   Minnesota prefers Wegovy and places Zepbound nonpreferred. Missouri and Wisconsin prefer Zepbound while treating Wegovy differently. North Carolina and Rhode Island prefer Wegovy and place Zepbound nonpreferred. Delaware and Kansas place both in the preferred column.

5. 5. Missing data is itself a policy finding.

   Delaware publishes product placement without a current public numeric threshold. Michigan's live detailed file crossed a version boundary. Missouri's PDL verifies preference but not the numeric clinical rule. Rhode Island's form collects BMI without printing a floor. Virginia's official sources describe conditional or conflicting frameworks. A clean-looking national table that fills those cells is more likely to be wrong than complete.

Which states changed GLP-1 obesity coverage in 2025 and 2026?

The change log shows why a single page-level “updated” date is not enough. Coverage ended, returned, expired, changed products, and crossed future-effective version boundaries within the same ten-month period. Each event needs its own effective date.

Source: Weight Loss Provider Guide Research change log, dataset cutoff July 31, 2026. Each row is tied to the state or original national-survey source in the primary-source register. Change sources: California; MassHealth Pharmacy Facts #276; Pennsylvania Medical Assistance Bulletin 2025112403; Utah current PA form; KFF January 2026 baseline.

How does the CMS BALANCE Model change state prior authorization rules?

BALANCE is a voluntary CMS Innovation Center model. For a participating state and a model drug, the State Agreement and supplemental rebate arrangement require a standardized access policy across fee-for-service and managed care. State applications were due July 31, 2026; a selected state must execute its State Agreement by January 1, 2027, and the model runs through December 2031.

The important word is participating. The Request for Applications does not make its criteria a nationwide Medicaid rule. A state with a more restrictive public policy is not violating BALANCE merely because the policy differs. The state first has to participate, and CMS can approve variations. CMS had not published a complete participant list at the cutoff, so this page does not label any state “noncompliant” and does not publish a count of state violations.

Source: CMS BALANCE Model State Medicaid Agency Request for Applications, §2.7.3, last modified March 2026. Separate attestation routes exist for type 2 diabetes, noncirrhotic MASH, and OSA. A participating state may adopt less restrictive criteria within the FDA-approved label.

Source: CMS BALANCE RFA §§2.4.2 and 2.7.3.

That lifestyle distinction matters. BALANCE requires provider attestation that lifestyle modification is current and will continue when clinically appropriate. It separately prohibits lifestyle-related step therapy. “Participating now” and “failed before approval” are not the same requirement.

Which GLP-1 weight-loss drugs does each state cover or prefer?

There is no national product answer hiding inside the state count. Product placement differs even among states with an active obesity pathway, and preference, coverage, and PA are separate facts.

Source: Current state PDLs, PA forms, and provider materials listed in the state evidence notes.

Why Ozempic is not an obesity-coverage proxy

Ozempic and Wegovy contain semaglutide, but they are different FDA-approved products with different labels. The current Ozempic label covers type 2 diabetes and specified cardiovascular and kidney-risk uses in adults with type 2 diabetes. It does not carry a chronic-weight-management indication.

Wegovy's current label covers chronic weight management, cardiovascular-risk reduction in specified adults, and noncirrhotic MASH with F2–F3 fibrosis. Zepbound's current label covers chronic weight management and moderate-to-severe OSA in adults with obesity.

Sources: Ozempic prescribing information; Wegovy prescribing information; Zepbound prescribing information.

Why can Medicaid cover the same molecule for one condition but not for obesity?

Federal law permits state Medicaid programs to exclude or restrict agents when used for anorexia, weight loss, or weight gain. The exclusion is tied to the use. That is why a product can have a covered medically accepted indication and a separate excluded weight-management use.

Under the Medicaid Drug Rebate Program, states generally cover participating manufacturers' covered outpatient drugs for medically accepted indications, subject to statutory exclusions and utilization controls. Section 1927(d)(2) of the Social Security Act, codified at 42 U.S.C. § 1396r-8(d)(2), expressly lists agents used for weight loss among the categories a state may exclude or restrict.

• Diabetes access does not prove obesity access.
• “The state dropped GLP-1s” is usually too broad. California, Massachusetts, and Pennsylvania ended obesity-use coverage while preserving separate pathways for other medically accepted indications.
• The prescribed indication changes the analysis. Wegovy cardiovascular or MASH treatment and Zepbound OSA treatment are not counted as obesity coverage on this page.
• Product labels matter. Two brands containing the same active ingredient can sit in different coverage pathways.

Children may be treated differently under EPSDT

Federal Early and Periodic Screening, Diagnostic, and Treatment rules require state Medicaid programs to provide coverable §1905(a) services found medically necessary to treat, correct, or reduce conditions for enrolled members under 21, even when the service is not included in the adult state plan. States determine medical necessity case by case.

That does not create automatic approval of a particular drug. It does mean the adult rule is not always the complete pediatric answer. Rhode Island's FFS PA04 form is not required for recipients under 21. North Carolina publishes a separate adolescent BMI pathway and a BMI-reduction renewal test. Kansas, Minnesota, Mississippi, and Tennessee publish age- or pediatric-specific treatment rules.

Source: CMS EPSDT guidance, especially “Other Necessary Health Care Services.”

Do Medicaid managed-care plans use the same criteria as fee-for-service?

Not automatically. A fee-for-service rule does not govern every Medicaid managed-care organization unless the state requires uniformity. The benefit channel is therefore a field in the dataset, not a footnote.

North Carolina's reinstatement bulletin applies to both Medicaid Direct and Managed Care. Tennessee administers the pharmacy benefit through TennCare managed care. Michigan operates a common-formulary framework across health plans. Rhode Island publishes an FFS-specific form and warns that managed-care rules may differ.

The practical verification sequence for an individual case is:

1. Identify the prescribed indication.
2. Identify whether the member is in FFS or an MCO.
3. Check the state policy.
4. Check the plan's current formulary and PA rule.
5. Treat a conflict as a live plan question, not as permission to average the documents.

BALANCE would remove this particular gap for participating states and model drugs because its terms require uniform criteria across FFS and managed care.

How was this dataset built?

We began with KFF's January 2026 survey of state Medicaid officials as a dated national baseline, then reviewed current public official materials through July 31, 2026. Each consequential field carries an evidence status. A blank cell is never silently converted into a rule.

Normalization rules

1. Indication first. A diabetes claim, cardiovascular pathway, OSA pathway, or MASH pathway never becomes obesity coverage by implication.
2. Channel declared. FFS is the national baseline. Managed-care differences remain visible.
3. Preferred, covered, and prior-authorized are separate fields. None is converted into another.
4. Inequality symbols are preserved. "Greater than 30" is not rewritten as "30 or greater."
5. Initial and renewal authorization are separate fields. A class default, titration edit, or claim quantity limit is not substituted for an initial clinical authorization.
6. A cutoff means a cutoff. Future-effective policy is logged and never backdated.
7. Archived and superseded sources are excluded from current fields. Delaware's archived form and Michigan's version boundary are explicit examples.
8. Missing means unfound. "Not publicly stated" does not mean no requirement exists.
9. Conflicts are disclosed. Virginia's official sources are not forced into a single number.
10. The national count is calculated from row status. It is not maintained as free-floating prose.

Source hierarchy

1. Current state Medicaid PDL or drug lookup
2. Current state PA form or clinical-criteria document
3. State Medicaid provider bulletin or manual
4. CMS-approved state-plan material
5. Enacted legislation or official budget language
6. Official MCO policy when the rule is plan-specific
7. Federal statute, CMS program material, and FDA labeling
8. Original national-survey producer for a dated baseline
9. Secondary reporting only for a clearly labeled implementation detail unavailable from the issuer

Relationship to prior peer-reviewed work

Klebanoff, Chetty, and Doshi published a cross-sectional study of state Medicaid anti-obesity GLP-1 policies using June 2024 data. The study identified 13 adult obesity policies, found two requiring at least two comorbidities, and reported that 70% specified qualifying comorbid conditions.

This July 2026 dataset is in the same manual-review tradition. It updates the policy date and adds product placement, effective-date control, initial authorization, renewal, channel, version status, and a 51-jurisdiction evidence ledger.

Source: Klebanoff MJ, Chetty AK, Doshi JA, Journal of General Internal Medicine, DOI 10.1007/s11606-025-10012-y.

What are the limitations of this data?

The principal limitation is update velocity. State Medicaid pharmacy policy can change between scheduled reviews, and agencies sometimes replace a file at the same URL without preserving the prior version. The second material limitation is equally important: 34 of the 51 jurisdictions carry a January 2026 national-survey baseline rather than an individual July state-source verification.

• The 34 baseline-only rows are presented as a dated survey result. They are not presented as freshly verified July negatives.
• Unpublished BALANCE participation. CMS accepted applications through the cutoff and had not published a complete participant list. A state decision not yet visible in public materials would not appear here.
• Incomplete current numeric criteria. Delaware, Michigan, Missouri, Rhode Island, and Virginia did not yield one current comparable numeric rule from the public sources used for this cutoff. Those are evidence-status fields, not invitations to fill the gaps from memory.
• FFS and managed-care divergence. FFS is the national comparison layer, not a guarantee that an MCO uses the same criteria.
• No approval or denial rates. We publish no state approval or denial rates because we did not locate a reproducible national public dataset that could support them.
• No individual coverage prediction. A published rule cannot account for every diagnosis code, claim edit, transition authorization, medical-necessity record, age rule, MCO variation, or exception process.

Frequently asked questions

Title:               Medicaid GLP-1 Prior Authorization Criteria by State (2026 Data)
Creator:             WPG Research Team
Publisher:           Weight Loss Provider Guide Research
Dataset cutoff:      2026-07-31
Last verified:       2026-07-31
Geographic coverage: United States — 50 states and the District of Columbia
Unit of observation: State Medicaid program, with FFS as the national baseline
Jurisdictions:       51
Active detail rows:  11
Evidence classes:    Current state-source review; documented 2026 change; KFF January baseline