How to Spot Fake Peptide COAs: 9 Checks (2026)

Learning how to spot fake peptides from a desk starts with the certificate of analysis (COA), because a document is far easier to fake than the material it describes. A fake COA is one of three things: a report the named lab never issued, a genuine report with edited fields, or a genuine report borrowed from a different product or lot. Nine checks catch most of them: the lab’s own record, the domain it sits on, uniqueness, lot match, dates, number precision, expected mass, internal arithmetic and a named method for every result. We ran all nine on the nine reports in our open lot report dataset, and the results are below.

Nine checks for a fake peptide COA at a glance

Source: Vortex Research Editorial Team. Check order follows the order in which each one can be run, from the lab record outward.

Research materials: Vortex Research publishes its third-party lab reports, each linked to the issuing lab’s own record, at /coa/. The values are also released as an open dataset, doi:10.5281/zenodo.23578844.

Three ways a peptide COA is faked

• Fabricated. The document was built from scratch, often on a template copied from a real lab. The named lab has no record of it. Check 1 catches this outright.
• Edited. A genuine report was changed: a purity figure raised, a date moved, a product name swapped. The lab holds a record, but its values differ from the copy. Checks 1, 5 and 6 catch this.
• Borrowed. A genuine, unedited report is presented for something it does not describe: a different product, fill or lot. Every value on it is real, and the portal confirms it. Checks 3, 4 and 7 catch this, and it is the hardest of the three to see.

The third type is why a matching portal record is a starting point rather than a verdict. A real report about the wrong lot is still the wrong report.

Checks 1 and 2: the lab’s own record and where it lives

A COA on a sales page is a copy. The original sits with the lab that issued it, and the lab is the only party the seller does not control. The full routine for opening a lab record and comparing it field by field is in our field-by-field guide to reading a peptide COA. Two points belong here because fakes exploit them.

First, the domain matters more than the link. A QR code or “verify” button can point anywhere, including a page the seller built to look like a lab portal. Reach the lab by typing its domain, or by finding it independently, and confirm the domain belongs to a laboratory with its own address and contact details.

Second, a lab that cannot find a report is an answer. Rule out a typing error, then contact the lab using details from the lab’s own website.

Check 3: recycled reports and reused traces

Borrowed reports show up when several documents from one seller are viewed side by side.

• Repeated report numbers. A lab assigns one number to one report. The same number on two products means at least one of them is borrowed.
• Identical chromatograms. Different peptides interact with a reversed-phase column differently and usually elute at different times. Two different compounds with the same retention time, the same peak shape and the same baseline noise are almost certainly one trace used twice.
• Reused vial photos. Some labs photograph the sample on receipt. The same photo, label angle and background under different lots is a copy.
• A report older than the product. A test date that predates the first listing of a product or size cannot describe current stock.

Checks 4 and 5: lot, fill and dates

The lot field ties a report to one production run, and lab reports such as those from ILS Laboratories state plainly that the submitter supplies it. A report is only useful if its material name, labelled fill and lot match the vial or the paperwork that came with it. A report for a 5 mg fill attached to a 10 mg product describes different material, even when the compound name matches.

Dates should run in a sensible order: received, then analysed, then reported. A report dated before its own received date has been edited or assembled. The test date should also be plausible for the stock being sold.

Checks 6, 7 and 8: impossible numbers

Some fakes fail on arithmetic alone.

Numbers outside what the method can produce

HPLC purity by area normalisation is a share of detected peak area (our HPLC purity explainer shows the calculation), so it cannot exceed 100%. A figure of exactly 100.00% means no other peak was integrated at all, which is unusual for a synthetic peptide carrying ordinary synthesis by-products such as deletion sequences and diastereomers [1]. Precision is the other tell. The reports in our dataset give purity to two decimal places, which is typical. A purity printed to four decimals, or one figure quoted to more places than every other figure on the same report, claims more resolution than the method delivers.

An identity mass that belongs to another molecule

Mass spectrometry confirms identity by comparing an observed mass with the mass calculated from the sequence. In electrospray ionisation, peptides pick up one or more protons, so one molecule shows up at several m/z values, one per charge state [2]. Those values can be worked out without an instrument. Using the monoisotopic masses in the Vortex peptide identity reference (doi:10.5281/zenodo.23578942):

Source: monoisotopic masses from the Vortex peptide identity reference; ion values computed for this article by adding the proton mass (1.00728 Da) per charge and dividing by the charge.

The TB-500 rows show why this matters. The name TB-500 is used for two different molecules: the acetylated heptapeptide Ac-LKKTETQ, near 889 g/mol, and full-length thymosin beta-4, a 43-residue peptide near 4963 g/mol. Vortex’s TB-500 is the full-length thymosin beta-4 material, and its product page lists that identity. The check is not which name is used but whether the measured mass matches the molecule the label and specification state. A thymosin beta-4 specification with an identity result near 4963 is consistent; the same document reporting a mass near 889, or a mass that matches neither, is the red flag. Our TB-500 vs thymosin beta-4 explainer sets out the two identities.

Parts that do not add up

Blend reports give a total and a per-component split. The split must sum to the total within rounding. A total that differs from its parts, or a split that repeats one component’s figure for both, points to a hand-edited line.

Check 9: missing methods and unverifiable testing claims

A result without a method cannot be checked or compared. A complete line names the test, the method and the result: purity by RP-HPLC at a stated wavelength, identity by LC-MS or by HPLC retention time against a reference standard, endotoxin by a named compendial method. Watch for “HPLC/MS” on a report that shows only a chromatogram, which leaves open whether mass spectrometry was run at all.

The same test applies to claims made away from a report. Statements on a product page such as “lab tested”, “HPLC and MS verified” or claims that cover a whole catalogue mean nothing without a report behind them that names the lab, the lot and the method. A test that appears in marketing copy but not on the report was not run, as far as the evidence shows.

We ran the nine checks on our own reports

The Vortex lot report dataset holds nine reports across eight materials (file generated 22 October 2026). It is a subset of the reports on /coa/. We ran the checks on it as a reader would: the published fields first, then each issuing lab’s own record.

Source: coa_reports.json, dataset generated 2026-10-22 (doi:10.5281/zenodo.23578844), compared with the ILS Laboratories portal records and the Freedom Diagnostics report file on 22 October 2026. Expected GHK ions computed from the identity reference.

Two results are weaknesses rather than passes. The MOTS-c lot field cannot be matched to a printed code, only to a cap colour. And eight of nine reports confirm identity by retention time, a valid but weaker test than a measured mass. Seven of the nine reports were issued to a partner facility rather than to Vortex, and are labelled that way on /coa/. None of this makes the reports fake. It marks where a reader has to rely on something other than the document.

What a clean result proves, and what it does not

A report that passes all nine checks is a genuine, internally consistent record of what one lab measured on one sample, under one lot code, on one date. Whether the material on the bench belongs to that lot is a separate question, answered by the lot code and the supplier’s records. It says nothing about lots without a report. For the wider set of supplier checks that sit around the document, see the checklist for evaluating a research peptide supplier.

Frequently asked questions

Can a genuine lab report still be the wrong report?

Yes. A report the lab really issued can be shown for a different product, fill or lot. The lab’s record will confirm it, because nothing on it was changed. The material name, labelled fill and lot code still have to match the vial or its paperwork.

What does an edited COA usually look like?

Its values differ from the lab’s own record of the same report number. On the document itself, signs include dates out of order, numbers with more decimal places than the rest of the report, fonts or alignment that change on one line, and results with no method named.

How can an identity mass be checked without a mass spectrometer?

Take the monoisotopic mass of the expected sequence from a reference source, add 1.00728 Da per proton and divide by the charge. For BPC-157 that gives about 1419.71 for the singly charged ion and 710.36 for the doubly charged ion. Observed values far from these describe a different molecule.

Is a COA image on a product page enough to rely on?

No. An image is a copy. It becomes evidence once the issuing lab’s own record confirms the same report number, lot, dates and results, and once the lot matches the material in question.

References

1. D’Hondt M, Bracke N, Taevernier L, et al. Related impurities in peptide medicines. J Pharm Biomed Anal. 2014;101:2-30. PMID 25044089. doi:10.1016/j.jpba.2014.06.012
2. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization for mass spectrometry of large biomolecules. Science. 1989;246(4926):64-71. PMID 2675315. doi:10.1126/science.2675315

Methods note: written by the Vortex Research Editorial Team. Dataset figures were computed directly from coa_reports.json (generated 2026-10-22). Expected ion values were computed from the monoisotopic masses in the Vortex peptide identity reference. Literature was selected from PubMed for analytical and quality-control topics, and every PMID was checked against its PubMed record for title, journal and year.

Vortex Research products are sold for in-vitro laboratory research by qualified researchers. Not for human or veterinary use. Not a drug, dietary supplement or cosmetic. Statements have not been evaluated by the FDA.