Military and veteran coverage of these drugs is built on the same kind of sentence that governs Medicare: not a rule about a molecule, but a rule about a purpose. The difference is that the two systems resolved that sentence in opposite directions, and one of them did it a decade ago. The Medicare version, its 2024 proposal and its 2025 reversal, is set out in the public-coverage article. This is the other one.
The statute says no
Title 10 of the U.S. Code sets the terms of the military health benefit, and its list of what may not be provided includes, in full: “Treatment of obesity may not be provided if obesity is the sole or major condition treated.”[1] The implementing regulation repeats it twice. The general exclusion list bars services and supplies related solely to obesity or weight reduction or weight control, whether surgical or nonsurgical, regardless of the circumstances under which performed.[2]
The second version is the one that reaches a prescription. Inside the morbid-obesity benefit, which otherwise authorizes bariatric surgery, sits an exclusion stating that payment may not be extended for weight control services, weight control or loss programs, dietary regimens and supplements, appetite suppressants and other medications, food or food supplements, exercise and exercise programs, or other programs and equipment primarily intended to control weight or for the purpose of weight reduction — regardless of the existence of co-morbid conditions.[2] That last clause is doing the same work as the Medicare exclusion: a comorbidity does not release it.
The regulation says yes, for two plans, through one kind of provider
And then the regulation turns around. Under the heading of benefits under TRICARE Prime and TRICARE Select, a paragraph reads that under the authority of 10 U.S.C. 1097 and sections 706 and 729 of the National Defense Authorization Act for Fiscal Year 2017, notwithstanding 10 U.S.C. 1079(a)(10), treatment of obesity is covered under TRICARE Prime and TRICARE Select even if it is the sole or major condition treated. It adds one condition: such services must be provided by a TRICARE network provider and be medically necessary and appropriate.[3]
That is a legislative override, written into the rule, sitting a few paragraphs from the exclusion it overrides. Medicare asked itself whether the phrase could be read differently and answered no; Congress answered the military version by amendment. The practical consequence is the network condition. A cash-pay telehealth service that does not bill insurance is, by construction, not a TRICARE network provider — so the benefit and the market described across the seller write-ups do not touch, whatever the statute says.
The pharmacy rule turns on one word
Medicines run through a separate regulation with its own uniform formulary, and it contains the sentence that decides whether a GLP-1 can appear on that list at all: pharmaceutical agents used exclusively in medical treatments or procedures expressly excluded from the benefit by statute or regulation will not be considered for the uniform formulary, and excluded agents are not available as non-formulary agents nor cost-shared under the pharmacy benefits program.[4]
Read the adverb. Semaglutide and tirzepatide are not used exclusively for weight reduction; they are diabetes drugs with a second approval, which is the distinction the two-indications article exists to draw. So the categorical bar does not catch them, and what remains is the ordinary formulary machinery: a Pharmacy and Therapeutics Committee selecting agents in each therapeutic class on relative clinical effectiveness and cost effectiveness, with the power to classify an agent as non-formulary or to impose preauthorization.[4] That is the ordinary apparatus of a drug plan, wearing a uniform.
What could not be read this session
The one thing a beneficiary most wants — the current uniform formulary tier and the prior-authorization criteria for each product — could not be retrieved. Every request to tricare.mil, health.mil and the TRICARE manuals site returned a JavaScript bot-protection interstitial rather than page content, and the Formulary Search Tool runs entirely in the browser with no reachable data endpoint. No figure on this page is taken from any of those sources, and nothing below should be read as the current formulary position. The authoritative answer is the Formulary Search Tool itself, opened in a browser.
What the claims data show the benefit amounts to
Utilization has been published, and it is modest. A cross-sectional analysis of the Military Health System Data Repository covering fiscal years 2023 and 2024 examined 568,232 adult TRICARE Prime and Plus beneficiaries aged 18 to 64 with a body mass index of 27 or higher and no diabetes. Of those, 31,176 — 5.5% — received at least one obesity medication.[5]
The composition is the interesting part. Users were 81% female and 71% dependents rather than serving members, split evenly between direct military care and the private sector. The medicines were led by phentermine at 41%, then semaglutide for weight management at 28%, naltrexone-bupropion at 20%, phentermine-topiramate at 16% and tirzepatide for weight management at 15%.[5] Adjusted odds of receiving anything at all were 0.64 for active-duty beneficiaries against others, and lower for Asian or Pacific Islander (0.59), Hispanic (0.81) and Black (0.82) beneficiaries.[5]
The authors note that TRICARE began covering obesity medications in 2018, that a policy change in September 2024 streamlined access, and that utilization remains low relative to disease prevalence, particularly among active-duty and minority beneficiaries.[5] The oldest and cheapest drug on the list is still the most prescribed one, which is a pattern worth holding next to the insurance-versus-cash comparison.
Among service members who do start, the molecule predicts who stays
A retrospective cohort in the same system followed 10,649 active-duty service members who started liraglutide, semaglutide or tirzepatide for weight management between January 1, 2021 and December 31, 2025. One-year persistence — time to a first gap in therapy beyond 90 days — ran 81.9% for tirzepatide, 70.7% for semaglutide and 34.2% for liraglutide (log-rank P < .001), with mean yearly proportion of days covered at 78.2%, 71.6% and 48.7% respectively.[6]
Against semaglutide, the adjusted hazard of discontinuation was 33% lower for tirzepatide (HR 0.67; 95% CI, 0.61 to 0.74), and male service members discontinued more often than female ones (HR 1.12; 1.02 to 1.22).[6] A daily injection performs worse than a weekly one on staying power in a population that deploys, trains and moves. That is an argument about dosing format rather than about efficacy, and it is the kind of difference a formulary decision made on clinical and cost effectiveness is not obliged to weigh.
The VA runs a different list entirely
Veterans Affairs publishes its national formulary as a machine-readable file, which means it can be read rather than paraphrased. The file retrieved on September 15, 2026 carries 5,678 product rows, twelve of which name a GLP-1 or dual incretin agonist.[7]
Two of the twelve are on the formulary. Ozempic (semaglutide injection) and Mounjaro (tirzepatide injection) both carry the flag the VA’s own application renders as Prior Authorization – Facility: on the national formulary, with requirements managed at the local facility level.[7]
The other ten are non-formulary, which in the VA’s own words means the product is not on the national formulary and requires a non-formulary drug request and prior approval to be dispensed. They are Wegovy and Wegovy HD, four separate Zepbound entries including the pen presentations, the semaglutide tablet, liraglutide (Saxenda and Victoza), dulaglutide (Trulicity), exenatide and the insulin-liraglutide combination.[7]
So the split is not between molecules. It is between the two diabetes-indicated injections that won their therapeutic class and everything else, including both weight-management brands, both liraglutide brands and the oral tablet. Non-formulary is a request rather than a refusal — but it is a request, with a decision at the end of it.
What the VA and DoD tell their own clinicians
The two departments write a joint clinical practice guideline, updated in 2025 on the basis of a systematic review of evidence published from April 2019 through January 2025, addressing twelve key questions and graded with GRADE. It recommends comprehensive lifestyle intervention as the foundation of care, found insufficient evidence to recommend for or against delaying medicines relative to that intervention, imposes no requirement to complete lifestyle treatment before starting anything else, and discourages stopping an effective medicine to prevent weight regain.[8]
That last clause matters more than it looks. A guideline that tells clinicians not to stop an effective drug in order to prevent regain sits awkwardly beside a formulary that puts most of the class behind an individual request, and both are official positions of the same department.
The readiness argument underneath all of it is stated plainly in the military literature: more than half of active-duty service members are classified as overweight and up to 27% have obesity, with obesity contributing to service disqualification, musculoskeletal injury and cardiometabolic disease at an estimated $2 billion a year in combined costs.[5] The Department of Defense instruction that sets the body-composition standards themselves could not be retrieved this session — the Executive Services Directorate returned an access error to every attempt — so no standard is quoted here.
Where the cash market meets this, and one thing our own reviews record
The population that falls through these rules buys retail, and that is where a specific friction shows up. Of the 296 seller reviews published here, a case-insensitive search of the review source files for the word TRICARE on September 15, 2026 returns exactly one: a telehealth service whose published terms state that patients with active Medicaid, Medicare or TRICARE coverage are not accepted at all, alongside a tier that covers the membership while the branded medication is billed separately.
That figure is a count of what our reviews record, and it is not a census of the market. The other 295 reviews have not been adjudicated on this question — most seller sites do not address it in a place a review would capture — so the right reading is that one company states the refusal plainly enough for it to be written down, and nothing at all is established about the rest. A beneficiary who needs to know has to ask the company. Separately, one seller in the roster describes itself as veteran-founded, which is a marketing fact about its ownership and says nothing about whether it accepts military coverage.
Two structural points follow. A TRICARE beneficiary who is refused by a cash seller has lost nothing they were entitled to, because the network condition in the regulation already put that purchase outside the benefit. And a beneficiary who pays cash anyway is usually buying a compounded preparation, which is not FDA-approved and is not reviewed by the FDA for safety, efficacy or quality before dispensing — a different product from the one the formularies above are arguing about.
What none of this settles
No statement here describes a particular person’s coverage. Plan type decides whether the obesity authority applies, the network condition decides which providers count, the uniform formulary decides which product and at what tier, and the VA decides non-formulary requests one at a time at a facility. Access also varies by more than the rules: veterans receiving primary care in the community had a 1.2 percentage point lower adjusted probability of a GLP-1 prescription than those in VA direct care, and Black and Asian veterans had 1.8 and 2.2 percentage point lower probabilities than White veterans across both settings.[9]
A denial under any of these systems can be appealed, and the levels and deadlines are in the appeals article.