Is Compounded Semaglutide Legal in 2026?

Short answer: compounded semaglutide was never made illegal by name. What happened instead is that the two legal doors it walked through both closed, and the small door that remains has a very tight frame.

If you want the practical version: the cheap, widely advertised, sign-up-online compounded semaglutide of 2023 and 2024 is effectively over. A narrow, genuinely patient-specific version still exists on paper, and some providers still sell something they say fits it.

Here is how the law actually works, and what changed.

What made compounded semaglutide legal in the first place?

Two provisions of the Federal Food, Drug, and Cosmetic Act allow pharmacies to make drugs without FDA approval.

Section 503A covers traditional compounding at a state-licensed pharmacy. A 503A pharmacy can compound a drug for an identified individual patient based on a valid prescription. It cannot compound, regularly or in inordinate amounts, products that are “essentially copies” of a commercially available drug [1].

Section 503B covers outsourcing facilities. These register with FDA, get inspected, and must follow current good manufacturing practice. In exchange, they can compound without a patient-specific prescription and ship office stock. But they can only use bulk drug substances that appear on FDA’s 503B bulks list, or where the compounded drug is on FDA’s shortage list at the time of compounding, distribution, and dispensing [1].

Semaglutide has never been on the 503B bulks list. The entire industrial-scale compounded GLP-1 business ran on the shortage clause. Semaglutide went onto FDA’s shortage list in 2022 as demand for Ozempic and Wegovy outran supply, and stayed there for nearly three years.

While it was listed, 503B facilities could legally make it from bulk powder, and FDA exercised enforcement discretion over 503A copies. That is the window telehealth companies drove through.

What closed the door: the shortage ended

On February 21, 2025 FDA declared the shortage of all doses of injectable semaglutide resolved, citing Novo Nordisk supply data showing that shipments met or exceeded U.S. demand [1]. Tirzepatide had already been delisted in December 2024.

FDA gave a wind-down period so patients would not be cut off mid-treatment: 503A pharmacies had until April 22, 2025 and 503B outsourcing facilities until May 22, 2025 to stop compounding, distributing, and dispensing semaglutide injection products [2].

The Outsourcing Facilities Association, the 503B trade group, sued in the Northern District of Texas to reverse the determination. It lost. The court denied a preliminary injunction on tirzepatide on March 5, 2025 and on semaglutide on April 24, 2025 [3], and FDA later prevailed on the merits, a result that was upheld on appeal [4].

After those dates, there was no shortage-based cover left.

What FDA said counts as “essentially a copy”

The obvious workaround was to make something that is not quite a copy. Add vitamin B12. Sell a dose that does not exist in the pen. Call it personalized.

On April 1, 2026 FDA spelled out where the line is, for the first time [1][5]. A compounded GLP-1 product is considered essentially a copy of a commercially available drug product if:

• it has the same active ingredient as the commercial product in the same, similar, or an easily substitutable strength, and
• the commercial product can be used by the same route of administration prescribed for the compounded product, regardless of how the commercial product is labeled.

The only escape is a prescriber determination. A compounded product is essentially a copy unless the prescriber determines and documents that it contains a change that produces a significant difference for an identified individual patient.

FDA then gave the example everyone had been waiting for. A compounded product combining semaglutide with another active ingredient such as vitamin B12 (cyanocobalamin) will be considered essentially a copy when the products are used by the same route, for example both injected, and the amounts of semaglutide and B12 in the compounded product are within 10 percent of the strengths of the respective commercially available products [5].

In other words, semaglutide plus B12 is not a get-out-of-jail card. It was the specific example FDA chose to rule out.

FDA added one piece of breathing room: it said it does not currently intend to take action against a 503A compounder for compounding an essentially-copy product regularly or in inordinate amounts if the compounder fills four or fewer prescriptions for that product in a calendar month [5]. That threshold is a compliance policy, not a safe harbor written into law, and four prescriptions a month is not a business model for a national telehealth platform.

One more caution from the lawyers who read the statement closely: FDA has not said what documentation a prescriber must create and keep to support a “significant difference” determination [5]. That ambiguity is where most of the remaining legal risk lives.

The 503B bulks list proposal: the door FDA is trying to weld shut

There was still a theoretical route back. If semaglutide were added to the 503B bulks list, outsourcing facilities could compound it from bulk even with no shortage. And if a future shortage occurred, the shortage clause would reopen on its own.

On April 30, 2026 FDA announced a proposal to close both possibilities [6]. It proposed not to include semaglutide, tirzepatide, or liraglutide on the 503B bulks list, after reviewing the nominations and finding no clinical need for outsourcing facilities to compound them from bulk drug substances. The FDA Commissioner framed it this way: when FDA-approved drugs are available, outsourcing facilities cannot lawfully compound using bulk drug substances unless there is a clear clinical need.

The formal notice published in the Federal Register on May 1, 2026 under docket FDA-2018-N-3240 [7]. Industry asked for more time to comment; on June 26, 2026 FDA extended the deadline from the end of June to July 30, 2026 [8].

As of September 14, 2026, the comment period has closed and FDA has not published a final determination [9]. Three things follow from that.

• Nothing changed on July 30. The proposal is not yet final.
• Semaglutide is already off both the shortage list and the bulks list, so 503B compounding is already unlawful today. The proposal would make that permanent rather than contingent.
• FDA said it will issue its response to Novo Nordisk’s related citizen petition at the same time as the final notice, so the two will land together [7].

The other pressure points

Even where compounding is technically possible, four other forces have squeezed it.

Ingredient imports. On September 5, 2025 FDA established Import Alert 66-80 and a “green list” of GLP-1 active ingredient manufacturers it has inspected or evaluated as compliant. Shipments from facilities not on the list are subject to detention without physical examination at the border [10]. That raised the cost and lowered the availability of the cheap overseas powder the compounded market ran on.

Marketing enforcement. On February 6, 2026 FDA announced it would move against GLP-1 active ingredients destined for non-approved compounded products and against misleading direct-to-consumer advertising, warning that failure to correct violations could bring seizure and injunction without further notice [11]. Three warning letter waves bracket that statement: more than 50 letters in September 2025; 30 letters to telehealth companies dated February 20, 2026 and announced by FDA on March 3, 2026 [17]; and 25 more posted on June 16, 2026 [18]. The stated violations are consistent across all three — claims implying sameness with an approved product, and branding a compounded drug with the telehealth firm’s own name in a way that implies the firm is the compounder.

Manufacturer litigation. As of an August 5, 2025 company statement, Novo Nordisk had filed 132 complaints in federal courts across 40 states and reported 44 permanent injunctions [15]. Eli Lilly sued four telehealth companies in April 2025 over “personalized” compounded tirzepatide marketing. Those cases have genuinely gone both ways: a judge in the Central District of California dismissed Lilly’s case against Willow Health Services with prejudice on February 3, 2026, while a judge in the Northern District of California let Lilly’s amended false-advertising case against Mochi Health proceed on April 20, 2026. The compounders’ own counterattack did not survive: Strive Specialties’ January 2026 antitrust suit against Lilly and Novo was dismissed on August 6, 2026 by Judge Micaela Alvarez in the Western District of Texas, who held that compounded and branded GLP-1s are not reasonably interchangeable and therefore do not form a single product market, and that “antitrust laws protect competition — not individual competitors” [16]. Either way, the cost of defending these cases is itself a deterrent.

State boards. State pharmacy, medical, and nursing boards are increasingly the sharp end. On August 19, 2026 Mississippi’s three licensing boards issued a joint statement barring licensed providers from compounding, administering, dispensing, advising, recommending, supplying, or prescribing non-FDA-approved or research-grade peptides, and from delegating those acts to others [12]. It is a board policy statement rather than a rule, but board policy is how licensure discipline gets imposed.

What about “research use only” peptides?

A separate market sells semaglutide and other peptides as powder labeled “for research purposes only” or “not for human consumption,” with no prescription.

FDA has rejected the labeling defense directly. In March 2026 it issued warning letters to Gram Peptides and Prime Sciences, concluding the products were intended as human drugs because of how they were marketed [13]. A further batch dated August 24, 2026 went to sellers listing retatrutide, semaglutide, tirzepatide, elamipretide, and tesamorelin next to bacteriostatic water for reconstitution. FDA has also stated that retatrutide and cagrilintide cannot lawfully be used in compounding at all.

The research-use-only label is a marketing convention, not a legal category that protects the seller.

One separate peptide development is worth not confusing with this one. At its July 23–24, 2026 meeting, FDA’s Pharmacy Compounding Advisory Committee voted to recommend six of seven peptides under review — BPC-157, KPV, TB-500, MOTS-c, epitalon, and semax — for the 503A bulks list, over the objections of FDA’s own scientific staff, and voted down the seventh, emideltide [19][20]. Those votes are recommendations only; they do not amend the 503A bulks list, and FDA has not acted on them. None of the six is a GLP-1, and nothing about that meeting changes the status of semaglutide, which was not on the agenda.

Is Congress going to change anything?

Possibly. The Safeguarding Americans from Fraudulent and Experimental (SAFE) Drugs Act of 2025, H.R. 6509, was introduced on December 9, 2025 by Representatives Rudy Yakym and Andre Carson, with a Senate companion from Senator Jim Banks [14]. It is not GLP-1-specific; it would rewrite compounding rules generally. Three provisions matter here [22]:

• It would codify “essentially a copy” in statute, defining it as any compounded drug containing an active ingredient found in a commercially available product, absent a change made for an individual patient that produces a significant difference. That is broader than FDA’s current definition, which also turns on dosage strength and route of administration.
• It would raise the copy-compounding frequency limit to 20 prescriptions a month, against the four-a-month threshold in FDA’s current enforcement policy. Read that carefully: on this point the bill is more permissive than the agency, not less.
• It would require a pharmacy, facility, or physician that compounds more than 20 times in a single month for patients residing outside its own state to file reports with federal regulators.

The bill was referred to the House Committee on Energy and Commerce on December 9, 2025 and remained at “Introduced” status with no floor action as of September 14, 2026. Treat it as a signal of where policy is being argued, not as current law.

So, is it legal right now?

Here is the honest summary as of September 14, 2026.

• Mass-market compounded semaglutide: no. Both the shortage pathway and the bulks list pathway are closed, and FDA is actively enforcing against the marketing.
• 503B outsourcing facility compounding of semaglutide: no. Semaglutide is on neither list.
• 503A patient-specific compounding: narrowly yes. A prescriber must determine and document a change producing a significant difference for that individual patient. Semaglutide plus B12 at standard strengths by the same route does not qualify.
• Research-use-only peptides sold without a prescription: no. FDA treats those sellers as marketing unapproved new drugs.
• Buying as a patient: not the target. Enforcement runs at sellers, not patients. That is not the same as saying the products are safe, and some state boards now discipline the providers who supply them.

If you are currently on a compounded product, this is a conversation to have with a healthcare provider rather than a decision to make alone. Branded self-pay prices fell substantially in late 2025 and 2026, and CMS’s Medicare GLP-1 Bridge Program began on July 1, 2026, giving eligible Part D beneficiaries a $50 monthly copay on Wegovy, Zepbound KwikPen, and Foundayo through December 31, 2027 [21] (verified 2026-09-14; the $50 is a program copay for eligible Medicare patients, not a cash price anyone can buy at). The arithmetic that pointed toward compounding two years ago may not point there now.

Sources

1. FDA clarifies policies for compounders as national GLP-1 supply begins to stabilize — FDA
2. FDA Resolves Semaglutide Shortage: Next Steps for Community Pharmacies, Telehealth Companies, and Other Providers — Alston & Bird
3. Semaglutide Compounding to End — American Med Spa Association
4. National Consumers League Welcomes the Appeals Court Ruling Backing FDA’s Decision to Remove GLP-1 Drugs from the Shortage List — National Consumers League
5. FDA Clarifies Policies for Pharmacy Compounders of GLP-1 Products — Foley & Lardner
6. FDA Proposes to Exclude Semaglutide, Tirzepatide, and Liraglutide on 503B Bulks List — FDA
7. List of Bulk Drug Substances for Which There Is a Clinical Need Under Section 503B — Federal Register, 91 FR 23431
8. Extension of Comment Period — Federal Register, June 26, 2026
9. Public Comment Closes on the FDA Plan to Permanently Bar Bulk Compounding of Semaglutide and Tirzepatide — Medical Daily
10. FDA Launches Green List to Protect Americans from Illegal Imported GLP-1 Drug Ingredients — FDA
11. FDA to Restrict Ingredients Used in Mass-Marketed Compounded GLP-1s, Crack Down on Misleading Ads — AJMC
12. Joint Statement Concerning the Prescribing of Non-FDA Approved (Research-Grade) Peptides, August 19, 2026 — Mississippi Board of Nursing, Board of Pharmacy, and State Board of Medical Licensure
13. Warning Letter: Gram Peptides, March 31, 2026 — FDA
14. Congress Introduces SAFE Drugs Act of 2025 — Dykema
15. Novo Nordisk expands legal action to protect US patients from unsafe, non-FDA-approved compounded “semaglutide” — Novo Nordisk, August 5, 2025
16. Lilly, Novo defeat antitrust lawsuit over GLP-1 drug market — Reuters, August 6, 2026
17. FDA Warns 30 Telehealth Companies Against Illegal Marketing of Compounded GLP-1s — FDA
18. US FDA sends 25 letters to telehealth companies over claims on compounded weight-loss drugs — Reuters, June 16, 2026
19. FDA Panel Backs 6 Peptides for Compounding — AJMC
20. FDA Advisory Committee Voted Yes on Six Peptides. Now What? — Buchanan Ingersoll & Rooney
21. Coming Soon: CMS to Provide $50 Monthly Access to GLP-1 Medications for Medicare Beneficiaries — CMS
22. H.R.6509 — SAFE Drugs Act of 2025, bill summary and status — Congress.gov