How to Read a Peptide COA and Spot a Fake

By Editorial team. Updated Oct 15, 2026. Awaiting editor sign-off

A certificate of analysis (COA) is a lab’s report on a sample. It is the main evidence most research-peptide vendors offer for quality, so it’s worth knowing how to read one and how to check it is real. A COA only helps you if three things hold:

1. The report is for the batch you are buying.
2. The lab exists, is independent of the vendor, and confirms it issued the report.
3. The tests on it measure what matters: identity, purity and quantity at a minimum.

This guide explains each part of a typical peptide COA, then shows how to verify one and what fakes tend to look like. For how individual test methods work and where they fall short, see our guide to peptide testing.

The anatomy of a peptide COA

Identity: is it the right molecule?

Mass spectrometry measures the mass of the molecules in the sample. A matching result means the main component has the expected mass. Labs list that expected value in their method details. One lab’s BPC-157 test page, for example, gives a molecular weight of 1419.5 along with the CAS number and amino-acid sequence, and lists HPLC as the method for its “ID, amount and purity analysis” [2]. So check which method a report used to confirm identity. If a COA reports purity but has no identity result, it hasn’t actually shown that the vial contains the peptide on the label.

Purity: how clean is it?

HPLC purity is the main peak’s share of everything the detector sees, given as a percentage. It does not tell you how many milligrams are in the vial. It also isn’t the same as peptide content. Freeze-dried peptides carry residual moisture and counter-ions such as trifluoroacetate and acetate, so only around 70–90% of the powder’s weight is typically peptide [6]. A highly pure peptide can still have a peptide content well below 100%.

Quantity: is the amount on the label in the vial?

Some labs report finished vials as “mg per vial and peptide purity” [2]. Quantity matters as much as purity, and the two can move independently. A 2024 study bought semaglutide from illegal online sellers and had it analyzed. Measured semaglutide content ran 28.56–38.69% above the label amount, while measured purity was only 7.7–14.37% against the 99% printed on the labels [7]. A COA that reports only one of these numbers leaves half the question open.

Optional tests

• Endotoxin. Usually measured with an LAL assay and reported as EU per vial or per ml. Check the stated range. One lab, for example, says its standard test is accurate up to 10 EU per vial and only estimates higher values [3].
• Microbial testing. Read the method name. Some labs sell “sterility testing” that is actually a total aerobic microbial count plus a total yeast and mould count (TAMC/TYMC), cultivation tests reported as pass or fail [4]. FDA guidance treats these as two different things: microbial enumeration is for non-sterile products, while sterility testing should use the pharmacopeial method, USP General Chapter <71> Sterility Tests [10]. Both are useful, but they aren’t interchangeable.
• Other add-ons. These include heavy metals (arsenic, cadmium, lead, mercury), net peptide content, residual trifluoroacetic acid (TFA), and variance testing across several vials from one batch [2].

How to verify a COA with the lab

1. Find the lab yourself. Search for the lab’s name and go to its own domain. Don’t follow links or QR codes printed on the PDF; a fake report can point to a fake lab page.
2. Use the lab’s verification route. Janoshik’s verification page, as archived in July 2025, asks for the task number, which is at the top left of each report, and a unique key printed at the bottom of the report. Public tests need no key [1]. Janoshik has also kept a public repository of tests that clients chose to make public since 2017 [5].
3. Compare every field. Check the client name, sample description, batch, dates, results and the chromatogram image against the lab’s copy. One mismatch is enough to treat the vendor’s copy as unreliable.
4. If the lab won’t or can’t confirm the report, treat it as unverified. That applies to in-house COAs with no third-party lab named.

A verified COA still only shows that the lab tested a sample. It doesn’t show that your vial came from the same batch. That is why independent blind purchases, where the buyer orders like any customer and sends the vial off for testing, carry more weight than reports a vendor commissions and selects. Our lab testing page explains how we’ll run blind purchases; we haven’t published results yet.

Signs of a fake or recycled COA

• The same chromatogram appears on different products or lots. Peaks, baseline noise and retention times differ from run to run, so identical images are a red flag.
• The batch number is missing, or it doesn’t match the vial.
• The dates don’t add up. The test is older than the batch, the report is years old, or a date is in the future.
• The lab can’t be found or contacted, or it refuses to confirm the report.
• The PDF shows visible editing: mismatched fonts, text out of alignment, or numbers pasted over a template.
• The report lists a test the lab doesn’t offer, or uses units that don’t fit the method.
• Every result is a flawless 99.9% across products, batches and years.

Paperwork alone doesn’t prove quality. In the Paradigm Peptides case, the owner was sentenced in July 2026 to 70 months in federal prison [9]. Regulated industries don’t trust supplier COAs blindly either. Under US dietary-supplement rules, a manufacturer must run its own identity test on each dietary ingredient. For other components, it may rely on a supplier’s COA only after first qualifying the supplier by confirming the COA’s results, and it must re-confirm them periodically [8].

How we use COAs

Our vendor profiles record whether each vendor publishes batch-specific COAs and names an independent lab. Both checks are part of our transparency checklist. That checklist covers public information only; it is not a purity test. For a list of vendors that publish batch COAs, and the labs named on them, see third-party tested vendors. Results from our own blind purchases, once published, will appear separately on our lab testing page.