How peptides are tested — individually, as a group, and what the labs actually check.

Every claim a vendor makes about a peptide — what it is, how pure it is, how much is in the vial — is just text on a label until an independent laboratory confirms it. On the gray market, where certification is inconsistent and sometimes absent entirely, that confirmation is the only thing standing between a research-grade compound and an expensive vial of the wrong powder. This article explains how compound testing actually works: what the labs look at, the levels of testing available, where most samples fall short, and how you can have a compound tested yourself — alone or together with others.

Why testing matters more on the gray market

Research peptides are sold "for research use only" and sit outside the regulatory framework that governs medicines. There is no agency inspecting batches, no mandatory release testing, and no legal accountability for a mislabeled vial. The certificate a vendor shows you may be genuine, outdated, generic, or fabricated — and from the outside, real and fake powder look identical.

The scale of the problem is not hypothetical. Janoshik, one of the most widely used analytical labs in this space, reported that a large share of the peptides it tested in 2024 failed to meet their claimed purity. Independent investigations have repeatedly found gray-market samples testing well below the advertised 99%, and in the worst cases containing a completely different compound — a vial sold as one peptide that mass spectrometry reveals to be another substance entirely, or in some cases no active peptide at all.

The levels of testing

"Tested" is not a single thing. A certificate of analysis can mean anything from a single purity number to a full pharmaceutical-style panel. Understanding the levels tells you how much a given COA actually proves.

The first two levels — identity and purity — are what a credible third-party COA should always contain together. Freedom Diagnostics, for instance, states the principle plainly: a purity number without identity confirmation is an incomplete answer, so they never issue one without the other. Content (assay) is the third pillar most serious labs add. Safety testing — endotoxins, microbial, heavy metals — is usually an add-on, and standard research-use COAs frequently omit it entirely.

Where most samples fall

The majority of gray-market COAs stop at level two: HPLC purity, sometimes paired with a mass-spec identity check. That covers "is it the right molecule" and "how clean is it," which is genuinely useful — but it says nothing about endotoxins, sterility, or heavy metals. Most research-use COAs do not include endotoxin or sterility testing unless the vendor specifically pays to add it.

This matters because the failures cluster in predictable places. GLP-1 analogues such as semaglutide, tirzepatide and retatrutide are among the most frequently mislabeled compounds, partly because a truncated fragment can pass an HPLC purity curve while carrying the wrong molecular weight — which only mass spectrometry will catch. The common failure modes, in roughly descending order of how often they appear in independent testing:

Testing individually — and as a group

You do not have to take a vendor's word for it. Anyone can send a sample to an independent lab and receive their own COA, tested blind — without reference to the supplier's claims — so the result is genuinely unbiased. Individual testing typically costs in the region of $75–200 for a standard identity-and-purity report, rising toward $200–1,000+ for a full panel with endotoxins, heavy metals and microbial analysis.

Because that cost is per sample rather than per vial, the community has long pooled it. Group testing — sometimes organized around a group buy — lets several people who bought from the same batch split the cost of a single independent test. One vial is sent in, one COA comes back, and everyone who shares that batch benefits from the result. Several labs support this directly: Janoshik, for example, offers an explicit group-testing route alongside individual orders. The economics are simple — a single test that would be expensive for one person becomes trivial split across ten, and the published result protects far more buyers than just the one who paid.

A newer model removes the cost entirely for the person submitting. Aggregator platforms like Finnrick test submitted samples for free and publish the outcomes openly, building a public dataset of vendor performance across thousands of samples — turning one person's test into a data point everyone can read.

The labs worth knowing

A handful of independent laboratories do most of the testing this community relies on. These are the names that appear most often on credible COAs and in community verification.

How to read what you get back

A COA is only as good as your ability to verify it. A static PDF is trivial to fabricate, so the single most important habit is to confirm the result on the lab's own server — via the verification key, QR code, or report ID — rather than trusting the document a vendor emails you. If a lab offers verification and the vendor cannot produce a working code, treat that as a red flag. Check that the report carries a specific batch number and a recent date; a COA more than twelve months old tells you nothing about the batch in your hand, because peptides degrade. And read all three of the numbers that matter together: identity (the right molecule), purity (how clean), and content (how much). A certificate that shows only one of them is, by definition, incomplete.