Peptide Certificate of Analysis: What It Proves and Does Not

A certificate of analysis (CoA) for a research peptide reports test results on one batch. The common headline figure, an HPLC purity percentage, describes the share of detected chromatographic signal that belongs to the main peak. It does not report how much peptide is in the vial, and it says nothing about endotoxin, sterility or metals unless those tests are listed. This article tests the general statements in our plain-language checklist against regulator documents and published studies.

Key points

• In the USP’s worked example, a peptide with 0.92% HPLC impurities still came to 0.93 mg of peptide per mg of material, because counter-ion and other weight-based components are counted separately [2].
• In the one published test purchase we found that measured purity and content, three semaglutide vials sold online showed purity of 7.7% to 14.37% against 99% claimed, and content 28.56% to 38.69% above the label [4].
• We found no published study that bought “research use only” peptides and measured their purity or content. The evidence below is about a human drug sold online, plus SARM products [4][5][6].
• A CoA from a lab is still the seller’s document. ILAC points to national accreditation bodies as the place to confirm a lab’s status and scope [9].

Does 99% purity mean 99% peptide in the vial?

No. Two different quantities are involved. The USP describes assigning purity by measuring every detectable impurity, “peptide related impurities, counter ion, water, residual solvents, non-combustible residues”, and subtracting them from 100% [2]. HPLC impurities are expressed as a percentage of total detected area; counter-ions and the rest are measured by weight [2]. Its theoretical example: 0.92% HPLC impurities, 5.58% w/w acetic acid and trace amounts of TFA and solvents give 0.93 mg peptide per mg of material [2]. An area-percent figure alone cannot produce that number.

Counter-ions are the charged partners that bind to a peptide’s charged residues; trifluoroacetic acid (TFA) is one of the most common [3]. A 2025 review says acetate has largely replaced TFA in peptide drug synthesis because TFA was found to affect stability and biological activity [3]. A lyophilised powder is therefore not pure peptide by weight. How much of it is peptide (often called net peptide content) is a separate assay, which the USP calls assaying “peptide mass content (mg/vial)” [2].

The semaglutide study shows the gap in practice. Its authors measured purity and content separately, and for the same three vials found low purity and content above the label [4]. The full text we read does not state how the purity percentage was calculated, so we cannot explain that combination [4].

Why do two labs report different purity?

The sources give one documented example, not a general rule. In a USP collaborative study of a desmopressin reference standard, ANOVA showed one of six laboratories differing significantly from the mean of the others. A different lab had not run water-content testing, so its data was removed from the value assignment [2]. Labs there also differed in within-lab variability and were weighted accordingly [2]. These were pharmacopoeial laboratories under a defined protocol. We did not find a study comparing vendor-commissioned labs, so reasons such as method choice or integration settings are not tested here.

What does mass spectrometry confirm?

Mass spectrometry measures mass. In the USP procedure, electrospray LC-MS/MS was used “to confirm the mass of the peptide and to verify the amino acid sequence” [2]; those are two aims. A CoA line giving only an observed mass speaks to the first aim. The sources we opened do not say whether a mass alone can establish the second.

A forensic example: a 2018 paper used LC-HRMS/MS and de novo sequencing on a seized injection vial and identified a previously undescribed heptapeptide, a glycine analogue of GHRP-2 (MW 874.02 Da) [7]. That is one seized vial, not a market survey.

Regulators do ask for more than a purity figure, in a different setting. FDA’s October 2017 draft guidance on synthetic peptide generics (glucagon, liraglutide, nesiritide, teriparatide, teduglutide) recommends orthogonal methods for primary sequence, secondary structure, aggregation and biological activity, and high-resolution methods such as UHPLC-HRMS for impurities [1]. It says applicants should identify each peptide-related impurity at 0.10% of the drug substance or greater, and proposes 0.5% as the ceiling for a new specified impurity, with characterisation and justification [1]. This is a draft for ANDA applicants, not a rule for research vendors. A search result describes a final version (May 2021); we did not open it. EMA’s synthetic-peptide guideline appeared in search results, but its server blocked us, so we cannot say what it requires.

What does a typical vendor CoA leave out?

We did not collect vendor CoAs, so we cannot say what a “typical” one contains. What the sources show is which tests are separate from HPLC.

• Endotoxin and sterility. The semaglutide study ran kinetic turbidimetric endotoxin tests per Ph. Eur. 2.6.14 and USP <85> and direct-inoculation sterility tests per Ph. Eur. 2.6.1 and USP <71>. No viable microorganisms were found, but endotoxin was detected at up to 8.9511 EU/mg [4]. A clean sterility result and a measurable endotoxin result coexisted in the same samples. FDA’s March 2026 Q&A guidance on pyrogen and endotoxin testing covers USP <85> methods for drugs and devices [10].
• Residual solvents and metals. FDA’s draft lists residual solvents, reagents and metals as “other impurities” a peptide drug can carry [1]. The USP example measures residual solvents and residue on ignition as separate tests [2].

“Lab tested” therefore tells you little until you know which tests, on which batch, by whom. We found no regulatory definition of “third-party tested” in the documents we opened.

Who issued it, and can that be checked?

ISO/IEC 17025 is the standard for testing and calibration labs. The ILAC MRA signatory search lists accreditation bodies by economy and scope (testing under ISO/IEC 17025, medical testing under ISO 15189, and others), and some entries link to a directory of accredited facilities [8]. ILAC says the accreditation body in the lab’s economy can confirm a lab’s accreditation status and scope, and that a counterfeit certificate should be reported to the body in the economy where it was issued [9].

Two limits follow. Accreditation covers a defined scope, so a lab accredited for one method is not automatically accredited for peptide purity. And accreditation says something about the lab, not about the sample: the seller chooses what is sent. We did not open any lab’s own report-verification page, so we cannot describe how those work. A report the seller commissioned remains the seller’s claim, and the checks in our checklist (batch number, date, named lab, raw chromatogram) concern the document, not the vial.

What do independent tests of online-bought products show?

Human drug product (semaglutide). Ashraf et al. (JMIR 2024) screened 1,080 search-result links and attempted test purchases from six sellers; three non-delivery scams meant three vials arrived, and all were rated probable substandard or falsified [4]. By LC-MS, content exceeded labels by 28.56% to 38.69%, with no peptide-like impurities identified, and purity was 7.7% to 14.37% against 99% claimed [4]. Funding came from the Hungarian Scientific Research Fund; two authors work for S-3 Research LLC, of which one is a cofounder [4]. The authors list the small, risk-based sample and that product quality may have changed in transit as limits [4].

Research peptides. We searched PubMed and Europe PMC on 2026-11-09 for purchase-and-test studies of peptides sold as research chemicals (BPC-157, tirzepatide, retatrutide and growth-hormone secretagogues among the terms). We found none reporting purity or content. That is a gap in what we found, not proof that no such study exists.

Related category (SARMs). Van Wagoner et al. (JAMA 2017) analysed 44 internet-sold products: 23 (52%) contained a SARM, 11 (25%) held an unlisted substance, and the labelled amount differed substantially from the measured amount in 26 (59%) [5]. A 2024 Italian study of 13 SARM products found the stated SARM in about 70%, and quantitative NMR gave 30% to 90% of the label claim; most sites used “for research only” wording [6]. These are small molecules, so they say little about peptide purity.

Regulatory context

“Research use only” is a label, not a legal category that exempts a product. In an FDA warning letter dated 2026-08-24 (ref. 735127) to a US seller of six products including semaglutide, retatrutide and tesamorelin, a footnote says the research-use and “not for human consumption” labelling does not change FDA’s finding that the products are intended for human use. The letter treats them as unapproved new drugs [11]. It does not mention CoAs, purity or endotoxin [11]. This covers the US only; we did not check the EU, UK or Germany.

Risks and unknowns

The semaglutide results come from three vials and cannot be generalised to other sellers or peptides [4]. We do not know how the purity percentage there was calculated. We have no study of research-use peptide batches, so how far vendor CoA figures match independent measurements is unknown. We did not verify how labs offer report verification.

What Peptilyze does and does not do

Peptilyze links to lab reports that shops publish. It does not verify them, and it does not rank shops by them. The shop page’s trust score includes a published-reports component, shown on that page. As of the API snapshot generated 2026-11-09 15:15 UTC, 34 shops are listed, and labReportShare is 0.156, meaning 15.6% of matched products carry a published lab report [12]. In the shop data, 19 shops have a labCoverage of 0, 13 have a value above 0, and 2 have none [13]. Shops that publish reports are collected at /lab-tests/.

Frequently asked questions

Does 99% purity mean 99% peptide in the vial? No. HPLC purity is an area percentage over detected peaks. Counter-ions, water and solvents are measured by weight and are not in that figure, so mass of peptide per mg of powder is lower [2].

How do I know a peptide CoA is real? A document alone cannot prove it. ILAC says the accreditation body in the lab’s economy can confirm status and scope [9]. Even then, the seller chose the sample.

What does third-party tested mean? We found no regulatory definition. It does not say which lab, which tests or which batch.

Does a CoA test for endotoxin or heavy metals? Only if those tests are listed. In the semaglutide study, endotoxin and sterility were measured separately from LC-MS purity [4].

What do independent tests of online-bought peptides show? For semaglutide vials, low measured purity and content above label in three samples [4]. For research peptides, we found no study.

Sources

1. FDA (CDER), “ANDAs for Certain Highly Purified Synthetic Peptide Drug Products That Refer to Listed Drugs of rDNA Origin”, draft guidance (October 2017). FDA PDF as mirrored at https://www.gmp-compliance.org/files/guidemgr/UCM578365.pdf (full text read; the fda.gov copy of this document could not be fetched, and the final 2021 version was not opened)
2. McCarthy D et al., Pharmaceutical Research, “Reference Standards to Support Quality of Synthetic Peptide Therapeutics” (2023). PMID 36949371, full text via Europe PMC (PMC10338602); funded by USP, whose staff are authors.
3. Liu Y et al., Biomolecules, “The Effects of Counter-Ions on Peptide Structure, Activity, and Applications” (2025). PMC12650639, full text via Europe PMC.
4. Ashraf AR et al., Journal of Medical Internet Research, “Multifactor Quality and Safety Analysis of Semaglutide Products Sold by Online Sellers Without a Prescription” (2024-11-07). PMID 39509151; abstract via Europe PMC and full text at https://pmc.ncbi.nlm.nih.gov/articles/PMC11582493/ (first 100,000 of 125,067 characters read).
5. Van Wagoner RM et al., JAMA, “Chemical Composition and Labeling of Substances Marketed as Selective Androgen Receptor Modulators and Sold via the Internet” (2017-11-01). PMID 29183075, via Europe PMC (abstract only).
6. Gaudiano MC et al., Sexual Medicine, “Illegal products containing selective androgen receptor modulators purchased online from Italy: health risks for consumers” (2024-03-27). PMID 38560649, via Europe PMC (abstract only).
7. Popławska M, Błażewicz A, Drug Testing and Analysis, “Identification of a novel growth hormone releasing peptide (a glycine analogue of GHRP-2) in a seized injection vial” (2018-08-14). PMID 30051972, via Europe PMC (abstract only).
8. ILAC, “ILAC MRA Signatory Search”. https://ilac.org/signatory-search/ (read 2026-11-09)
9. ILAC, “Accredited Conformity Assessment Bodies”. https://ilac.org/ilac-mra-and-signatories/accredited-conformity-assessment-bodies/ (read 2026-11-09)
10. FDA, “Pyrogen and Endotoxins Testing: Questions and Answers” (revised March 2026; page current as of 2026-03-18). https://www.fda.gov/regulatory-information/search-fda-guidance-documents/pyrogen-and-endotoxins-testing-questions-and-answers (guidance page only; the document itself was not downloaded)
11. FDA, Warning Letter to Peak Performance Peptides, ref. 735127 (2026-08-24). https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/peak-performance-peptides-735127-08242026
12. Peptilyze, public API /market (generatedAt 2026-11-09T15:15:23.440Z), read 2026-11-09. https://peptilyze.com/api/v1/market
13. Peptilyze, public API /shops (generatedAt 2026-11-09T15:15:23.440Z), read 2026-11-09. https://peptilyze.com/api/v1/shops

This article is information, not medical advice.