Testing and certificates of analysis: what they prove, and what they don’t

What the number on the certificate means

A purity assay, usually high-performance liquid chromatography, separates the components of a sample and reports the share of the total that corresponds to the target peptide. That is a genuine measurement, and a low result is genuinely informative: ECRI and ISMP, reviewing analytic testing of gray-market peptide products in April 2026, reported purity ranging from 5% to 75% of label, with arsenic and lead above the thresholds accepted for injectable drugs — though the white paper does not identify the testing laboratory, which is itself worth noticing on a page about what a certificate proves. Eli Lilly, announcing an International Trade Commission ruling in December 2024, said that almost half of the knockoff tirzepatide it had tested came in below the potency stated on the label, that some contained no tirzepatide at all, and that it kept finding bacteria, high endotoxin and high impurity levels; in March 2026 it flagged an impurity specific to preparations combining tirzepatide with vitamin B12. Testing found all of that. The point of this page is not that testing is worthless — it is that a purity percentage answers one question out of the several that matter, and the other questions do not become answered by being left out of the report.

Four things a purity result does not settle

Sterility. Purity and sterility are different assays. A powder can be 99% the right peptide and carry bacterial contamination; nothing in a chromatogram would show it. For a substance intended to be reconstituted and injected, this is the omission that matters most.

Endotoxins. Bacterial endotoxins survive processes that kill the bacteria themselves and cause a febrile reaction when injected. Testing for them is a separate assay again, and one rarely present in the paperwork circulated with gray-market vials.

What the rest of the vial is. A result of “98% pure” does not name the other 2%. Peptide synthesis leaves behind truncated and deleted sequences, residual solvents and reagents; some are inert, some are not, and a purity figure treats them all as a single anonymous remainder. Heavy metals — the finding that made the ECRI and ISMP review notable — require yet another assay.

Provenance and chain of custody. A certificate describes a sample, not a batch and not a supplier. Nothing in it establishes that the vial delivered came from the batch tested, that the material was made under controlled conditions, or that storage and transport preserved it. In the regulated system those assurances come from manufacturing standards and inspection, not from an analytical report.

Multi-peptide blends compound the problem in a specific way. In a vial sold as GLOW or KLOW, a single purity figure cannot tell you whether four peptides are present in the stated ratio, or whether one of them is present at all. The question the buyer thinks the certificate answers is not the question the assay was run on.

Testing does not change legal status — anywhere

This is the part that gets lost. In the United States, whether a substance may be compounded under section 503A turns on whether it is the subject of a USP monograph, a component of an approved drug, or on the 503A bulks list. Analytical results from a private laboratory are not one of those routes. A certificate of analysis does appear in the statute — but as a requirement attached to bulk substances that already qualify, alongside the requirement that the substance be manufactured by an establishment registered with the FDA. That is the regulated version of testing: a certificate from an inspected manufacturer, for a substance that is already permitted, verified again by the pharmacy that uses it. A report commissioned by a seller of an unlisted substance is a different object performing a different function.

German law makes the same distinction more explicitly. A pharmacy compounding for an individual patient may only use starting materials whose pharmaceutical quality is established: § 11 ApBetrO requires the pharmacy to carry out its own identity test on the material in front of it, and a manufacturer’s test certificate can substitute for the wider analysis only where that manufacturer holds a manufacturing authorisation under § 13 AMG. A chromatogram from a third-party lab, commissioned by a research-chemical vendor, satisfies neither limb. And for the peptides on the FDA’s agendas there is no pharmacopoeia monograph to test identity against — the reference the whole procedure depends on does not exist. See our Europe and DACH page for how that plays out across Germany, Austria and Switzerland.

Why a testing industry grew anyway

Demand for these tests is real and enormous. Bloomberg reported in June 2026 on the analytical laboratory outside Prague, Janoshik Analytical, that built its reputation checking black-market steroids and now tests peptides at scale: its peptide test volume rose more than 1,200% between 2023 and 2025. Read that number carefully. It is a measurement of how much unregulated injectable material is moving through private hands, and of how little the people holding it trust their suppliers. A functioning regulated supply does not generate that demand curve, because identity, purity, sterility and provenance are settled upstream by manufacturing standards and inspection rather than downstream by the person about to inject. The growth of third-party testing measures the size of the problem. It is not the solution to it, and no laboratory claims otherwise.

A related pattern is worth naming: a testing badge on a seller’s website is marketing. The laboratory tested a sample somebody sent it and reported a number; it did not inspect a facility, approve a product, or vouch for the vial that arrives. Presenting an analytical result as a mark of legitimacy inverts what the result is.

What this page is not

It is not a guide to vetting a supplier, reading a chromatogram, or choosing a laboratory. That would be sourcing advice wearing a lab coat, and it would rest on a premise this site does not accept: that a sufficiently careful buyer can make an unauthorised injectable safe. The variables that determine whether an injectable is safe — sterile manufacture, controlled conditions, verified identity against a recognised standard, an intact chain of custody — are properties of a supply chain, not of a document. Where that chain exists, the testing is already built into it and nobody has to commission anything. Where it does not, a certificate cannot construct one after the fact.

Primary sources: ECRI and ISMP, “White Paper on Compounded Peptide Products” (© April 2026, announced 5 May 2026) — the purity range is reported there without attribution to a named testing laboratory · Eli Lilly press release of 23 December 2024 on the International Trade Commission ruling, reporting its testing of knockoff tirzepatide, and Lilly’s notice of March 2026 on a novel impurity in tirzepatide/vitamin B12 combinations · section 503A of the FD&C Act on bulk drug substances, including the requirement that a substance be accompanied by a valid certificate of analysis and be manufactured by an establishment registered with the FDA · Apothekenbetriebsordnung (ApBetrO) §§ 6 and 11 on starting materials and identity testing, and Arzneimittelgesetz (AMG) § 13 on manufacturing authorisation · European Pharmacopoeia general requirements for substances for pharmaceutical use · Bloomberg reporting of 17 June 2026 on the peptide market, for the scale of gray-market test demand. This site does not link to or evaluate sellers or testing services.