Peptides

Peptide safety in 2026: the three legal channels, what testing finds in gray-market vials, and how to read a certificate of analysis

By the US Health Digest editorial team · Published August 10, 2026 · Every claim linked to its primary source

Every peptide for sale in the United States sits in one of three legal channels: an FDA-approved drug, a compounded drug made under the bulks-list rules of sections 503A and 503B, or a gray-market vial sold "for research use only" — which FDA treats as an unapproved new drug regardless of the label. In a March 2026 warning letter to a peptide vendor, FDA wrote that despite "Research Use Only" labeling, "evidence obtained from your website establishes that your products are intended to be drugs for human use." Published test purchases show why the channel matters: in a 2024 peer-reviewed study, every semaglutide vial delivered from illegal online sellers contained bacterial endotoxin and 28.56% to 38.69% more drug than labeled. A certificate of analysis is a vendor-supplied batch document, not a license: under 21 U.S.C. § 355, no new drug may be introduced into interstate commerce without an approved application, and no lab printout changes that. In FDA's own review of BPC-157, the certificates in both nomination packages did not match the substance nominated.

The three channels a peptide can legally exist in

"Peptide" is a chemistry word, not a legal category — one word covering three different markets. Which rules apply depends entirely on the channel a given vial travels through.

Channel 1: FDA-approved drugs. Semaglutide, tirzepatide, tesamorelin and insulin are all peptides with approved labels, made under federal manufacturing standards and dispensed by prescription — the GLP-1 drugs are the proof that peptides can clear the full bar. Approval attaches to a specific product from a specific manufacturer, not to the molecule.

Channel 2: state-licensed compounding under sections 503A and 503B. A compounding pharmacy may use a bulk peptide only if it has a USP/NF monograph, is a component of an approved drug, or appears on FDA's 503A bulks list — a list built by notice-and-comment rulemaking, created by a 2019 final rule. A substance that clears none of those three gates leaves the compounded product — in the words of FDA's January 2025 interim policy — "not eligible for the exemptions in section 503A and may violate the FD&C Act." While nominations wait, FDA sorts substances into interim categories; Category 2 holds those with "significant safety risks." Two traps here. First, the 503A and 503B interim policies are separate tracks: FDA's category page notes that ipamorelin acetate sits both among withdrawn nominations and in Category 2 under the 503B policy — withdrawing under one track leaves the listing under the other standing, a trap we documented in our CJC-1295/ipamorelin report. Second, an advisory vote is not a rule: the Pharmacy Compounding Advisory Committee's July 23–24, 2026 recommendations on seven peptides were non-binding, and as of August 10, 2026, our search of the Federal Register found no proposed or final rule adding any of them — searching "BPC-157" returns only the meeting notice. Our BPC-157 report covers why even a winning vote changes nothing by itself.

Channel 3: "research chemicals." Everything else — the vials sold online with a disclaimer that they are "not for human consumption" — is outside both channels. FDA told a compounder in 2019 that BPC-157 meets none of the 503A criteria, and told the vendor in the March 2026 letter that its injectable peptides were unapproved new drugs whose route of delivery "bypass[es] some of the body's key defenses against toxins and microorganisms." "Research use only" is a liability disclaimer aimed at the seller's exposure, not a safety statement about the vial.

What published testing has found in gray-market vials

The peer-reviewed record on this supply chain is small but consistent. The most direct evidence is the 2024 study in the Journal of Medical Internet Research: researchers surveyed online sellers offering semaglutide without a prescription, then made test purchases from six illegal online pharmacies. Three pre-filled pens never arrived — nondelivery scams — and the three vials that did arrive all failed: semaglutide content exceeded the label by 28.56% to 38.69%, endotoxin was detected in every sample (2.1645 to 8.9511 EU/mg), and measured purity of the lyophilized samples ranged from 7.7% to 14.37% — against the 99% claimed on the labels.

That pattern is not new. Belgium's official medicines control laboratory, which analyzes peptide products seized by its medicines agency and customs, built its screening method around 25 different peptides — substances already found in illegal preparations seized across Europe, plus others not yet confiscated but already discussed on user forums — including some still in preclinical or clinical development. And the problem reaches even the legitimate research-supply chain: a 2020 analysis co-authored by scientists at the United States Pharmacopeia found 20% and 43% by weight of undeclared mannitol in two commercially sourced custom synthetic peptides — an adulterant the vendors' routine quality control had missed, in products that, as the authors note, "are usually trusted rather than being considered flawed." A 2025 commentary in the Annals of Pharmacotherapy describes the consumer-facing version: semaglutide and tirzepatide sold "for research purposes only" to buyers who often lack the supplies and knowledge to reconstitute and dose them.

For the most popular gray-market peptides specifically, the gap runs the other way: we found no published, peer-reviewed purity survey of retail BPC-157 products — so no independently verified figure exists for what an average vial contains. What is documented sits in FDA's own BPC-157 briefing document: peptides aggregate and degrade, aggregates and impurities drive immunogenicity, and injected BPC-157 "may pose a significant risk for immunogenicity."

What a certificate of analysis is — and what it cannot prove

A certificate of analysis (COA) is a document reporting laboratory test results for one batch of material. A meaningful one covers four things, and each maps to a specific test: identity — mass spectrometry confirming the molecular weight and sequence match the named peptide, the method the Belgian control laboratory uses for exactly this purpose; purity — a chromatographic (HPLC) percentage of the peptide versus its related impurities; quantity — how many milligrams are actually in the vial, the number the JMIR test purchases found off by roughly a third; and, for anything injected, sterility and endotoxin — the same study found endotoxin in every vial it tested. FDA's briefing document adds a subtler failure mode: identity has a salt-form dimension, and in both BPC-157 nomination packages FDA evaluated, the COA described the acetate while the identifying codes matched the free base.

Now the limits, which matter more. A COA describes a batch, not your vial — nothing ties the printout to the unit shipped to you. It is supplied by the seller, is frequently generated by an unaccredited lab the seller chose, and is trivial to reproduce as an image. High purity is not safety: a vial can be 99% pure peptide and still carry endotoxin, be the wrong dose, or be a molecule with no adequate human safety data at any dose. And a COA has no legal force. Under 21 U.S.C. § 355, introducing a new drug without an approved application is prohibited — a lab report cannot make an unapproved drug legal to sell for human use, which is why FDA's warning letters turn on what the product is intended for, not on what a vendor's paperwork claims it contains.

COA red flags

Questions to ask a prescriber offering peptide therapy

There are legal channels for peptide medicine: FDA-approved products, and properly licensed compounding within the bulks-list rules. These questions sort one from the other.

• Is this an FDA-approved product? If not, which channel is it — a state-licensed 503A pharmacy or a registered 503B outsourcing facility — and can you name it?
• Is the bulk substance on the 503A bulks list, in an approved drug, or the subject of a USP monograph — or is it in Category 2, which FDA associates with "significant safety risks"?
• May I see the certificate of analysis — with the lot number, the testing lab's name and accreditation, and endotoxin and sterility results if this is injected?
• What human trials support this peptide for my goal? (For the ones we have audited, the answer was thin to nearly empty.)
• What does it cost, and what would an approved alternative cost? Our peptide-therapy cost review is a starting point.

A prescriber working in the legal channels can answer all five without flinching. A vendor whose answer is a screenshot of a certificate is telling you which channel you are in.