Where research peptides come from

Three completely different regulatory positions arrive in near-identical vials. Learn to tell them apart upstream, because the label is the last place that will tell you.

Last updated 22 August 2026

Three supply channels and what governs each

PC-ART-004 Reference 22 August 2026 7 min read

An opened cardboard shipping box on a kitchen counter, packing material pushed aside, a small unlabelled carton visible inside. — A parcel tells you nothing about which of three regulatory positions its contents occupy. That was decided a long way upstream, and it is rarely printed.

Approved product

Three channels feed this market, and telling them apart is the single most useful thing you can learn about sourcing. Start at the top. An approved peptide has finished trials, holds a marketing authorisation, and is made under current good manufacturing practice. Semaglutide and tirzepatide are the household examples: semaglutide holds separate applications for Ozempic, Rybelsus and Wegovy, tirzepatide for Mounjaro and Zepbound, each approved on its own date for its own indication.

Now the detail that does most of the work, because it is the one everybody blurs. Approval attaches to a specific product — a named formulation, from a named manufacturer, for a named indication. Not to a molecule. "Semaglutide is FDA-approved" is true of Novo Nordisk's products, and it is not a property the molecule carries with it into a vial somebody else filled.

What approval covers

A specific product from a specific manufacturer, evaluated for safety, effectiveness and manufacturing quality against an indication.

Identity and purity are a regulatory requirement with a specification
Storage and dosing are published on the label
Adverse events are collected and reportable

What it does not cover

Anything else containing the same molecule. The approval does not travel with the chemistry.

Compounded versions of the same molecule are not approved
A research vial of the same peptide has no approval of any kind
An off-label use is not covered by the indication that was approved

Compounded preparation

The middle channel is the one people find genuinely surprising. Compounding pharmacies work under sections 503A and 503B of the Federal Food, Drug, and Cosmetic Act, dispensing against a prescription rather than under an approval — and the exemption from premarket approval is the whole point of the arrangement. A compounded drug is not evaluated by the FDA for safety, effectiveness or manufacturing quality. "Pharmacy-grade" sounds like a standard and is not a regulatory term at all.

The statute is far more precise than its reputation, and this is worth committing to memory: there are three routes by which a bulk substance can be compounded, and most summaries give you two. It qualifies if it complies with an applicable USP or National Formulary monograph where one exists; or, with no monograph, if it is a component of an approved drug; or, with neither, if it appears on the FDA's 503A bulks list. Meet none of the three and it may not be used. That is a determinate answer rather than a grey area — and it excludes most of the peptides this market talks about.

Their position moved in 2026, and the direction is very easy to misread in your own favour. BPC-157, TB-500, CJC-1295, ipamorelin and injectable GHK-Cu all sat in Category 2: substances nominated for compounding that the FDA identified as presenting significant safety risks. The nominators withdrew those nominations, which took them out of Category 2. It did not put them on the 503A bulks list. Being out of a risk category is not the same as being allowed.

Rows of capped glass vials standing on wire shelving inside a lit refrigerated cabinet. — Cold storage looks identical whichever channel filled the vials. Nothing visible in this room separates an approved product from a compounded one.

Three channels, and what each is evaluated against

4 categories · 8 compounds

Channel	Records	Evaluated for	Not evaluated for
Approved product	Semaglutide, Tirzepatide	Safety, effectiveness and manufacturing quality, per indication	Uses outside the approved indication
Compounded preparation	Tesamorelin	State pharmacy practice; the substance must clear one of three statutory routes	Safety, effectiveness or manufacturing quality — exempt by design
Research use only	BPC-157, TB-500, Ipamorelin, GHK-Cu	Nothing. No purity specification applies	Identity, purity, sterility, endotoxin, fill accuracy, or content
Investigational, in trials	Retatrutide	Trial protocols and their oversight, inside the trial	Any supply outside a trial, which has no authorised route at all

Research use only is a seller's phrase

Which brings us to the third channel and the six words holding it up. "For Research Use Only. Not for use in diagnostic procedures." is a labelling statement defined in US regulation for in vitro diagnostic products — devices. There is no pathway anywhere that makes a bulk injectable peptide lawful to sell because a label carries that sentence. What governs is intended use: a product intended to affect the structure or function of the body is a drug, and putting an unapproved new drug into interstate commerce is a prohibited act. The FDA has sent warning letters to sellers on exactly that basis.

So it is a disclaimer, not a classification — and that is also why nothing else printed on the vial can be leaned on. Material outside the drug framework carries no purity specification, no identity requirement, and no sterility or endotoxin standard, because none of those attach to a product no regulator is evaluating.

The quality argument you sometimes hear runs backwards as well. Skipping the approval pipeline gets described as what makes high-purity material available; purity is precisely the thing it cannot deliver. The FDA's stated concerns for these substances are peptide-related impurities, aggregation, immunogenicity and difficulty of characterisation — which is, exactly, the list a specification would exist to control.

Intended use governs
What a product is for is determined by how it is presented and sold, not by a sentence disclaiming it. A vial marketed with dosing information for people is being offered as a drug whatever the label says.
A certificate is not oversight
A buyer-obtained certificate of analysis is not a quality-assurance system. It is a document supplied by or on behalf of the seller, and nothing external verifies that it corresponds to the vial that arrived.
Named absence, not silence
No census, registry or trade body counts this market, because it is unregistered. Any figure quoted for vendor numbers, failure rates or contamination rates is an estimate someone made, and is usually printed without saying so.

What is measured, and what is asserted

Finish with the scoreboard, because two of these failures are measured and the rest are folklore, and the difference matters more than the numbers. Fill accuracy against label claim is measured by independent testers, underfill and overfill both count as failures, and vendors do fail. Purity against label claim is measured too — the sharpest figure published remains that 2024 study of no-prescription online semaglutide, where three of six orders arrived, all three were labelled 99% purity, and all three assayed between 7.70% and 14.37%.

Heavy-metal contamination is the standard third item on every list of this kind, and nobody is measuring it. No published dataset of independent peptide testing reports heavy-metal results at scale. A claim that testing routinely exposes toxic metals is describing a test that is not being run rather than a result somebody found, and it does not belong in the same sentence as the two above.

Domestic versus overseas belongs in that same category. Nothing we found compares product quality, seizure rates or supply failure between US and offshore sellers — and the comparison fuses two unlike things anyway, since domestic compounding is a licensed activity requiring a prescription, while a domestic research-chemical seller sits under exactly the same absence of oversight as an overseas one.

503A: The section under which a pharmacy compounds for an identified patient on a prescription. Exempt from premarket approval; the bulk substance must clear one of three statutory routes.
503B: An outsourcing facility, which may compound without a patient-specific prescription and registers with the FDA. Also exempt from premarket approval.
Bulks list: The FDA's list of bulk drug substances that may be used in compounding where no monograph exists and the substance is not a component of an approved drug.
Category 2: Substances nominated for compounding that the FDA identified as presenting significant safety risks. Withdrawal of a nomination removes a substance from the category without adding it to the bulks list.
Research use only: A labelling statement defined for in vitro diagnostic products. Applied to an injectable peptide it is a seller's disclaimer with no regulatory effect.
Unapproved new drug: A product intended to diagnose, treat or affect the structure or function of the body without an approved application. Introducing one into interstate commerce is a prohibited act.

Sources and evidence

11 sources

U.S. Food and Drug Administration. Drugs@FDA application records for semaglutide and tirzepatideRegulatory database