How to Read a Peptide Certificate of Analysis (COA)

What a COA actually is — and what it can never be

A Certificate of Analysis (COA) is a batch-specific lab report. It documents the analytical tests a manufacturer (or a third-party lab) ran on one particular lot of material, and it lists the results against a set of pre-defined specifications. A typical peptide COA names the product and sequence, gives a batch or lot number and a manufacturing or test date, states a molecular formula and theoretical molecular weight, and reports the outcomes of techniques such as high-performance liquid chromatography (HPLC), mass spectrometry (MS), and — sometimes — water content, counterion content, endotoxin level, and sterility.

Here is the part most buyers get wrong: a COA describes a sample, not necessarily the vial in your hand. The strength of any COA rests on three assumptions that the document itself cannot prove — that the testing lab is real and competent, that the data was not edited in a PDF, and that the batch number on the certificate genuinely matches the batch number on your vial. A COA is evidence about chemistry. It is not evidence of safety, of regulatory approval, or of fitness for any use in a human or animal. The overwhelming majority of peptides sold with COAs in the gray market are unapproved research chemicals, and no certificate changes that status. Read a COA the way an auditor reads a financial statement: useful, structured, and worth exactly nothing if the underlying numbers are fabricated.

HPLC purity: the headline number, and its blind spots

The figure most people fixate on is HPLC purity — usually a percentage like 98% or 99%. Reversed-phase HPLC separates the components of a sample as they pass through a column, and a UV detector measures absorbance, typically in the 210–220 nm range where the peptide bond absorbs. Each separated component shows up as a peak. Purity is reported as the area of the main (target) peak divided by the total area of all detected peaks. A '99% purity' result therefore means the target peptide accounts for 99% of the UV-absorbing, chromatographically separated species, and the remaining 1% is made up of impurities such as truncated sequences, deletion products, or degradation fragments.

That definition carries two important blind spots. First, HPLC-UV purity is relative, not absolute: it tells you the target is the dominant peak among what the method can see, not how much actual peptide is in the vial. Water, salts, and counterions are largely invisible to this measurement. As a 2025 paper in Pharmaceuticals (PMC12389442) put it plainly, standard HPLC-UV purity 'fails to reveal counterion composition' — a peptide carrying a heavy load of trifluoroacetate can look just as 'pure' as one that has been counterion-exchanged. Second, purity is method-dependent. A peak that co-elutes (overlaps) with the target can hide inside the main peak and inflate the number. This is why a credible COA shows the actual chromatogram — the trace with peaks and retention times — and not just a bare percentage. A purity figure with no chromatogram attached is an assertion, not a measurement.

Mass spectrometry: identity, not quantity

If HPLC answers 'how much of the sample is the main thing,' mass spectrometry answers 'is the main thing the right thing.' MS measures the mass-to-charge ratio of ionized molecules, allowing the lab to determine the molecular weight of the compound and compare it against the theoretical mass calculated from the stated sequence. A COA should list both an expected (theoretical) mass and an observed mass; when they agree within the method's tolerance, that is evidence the synthesized molecule has the intended molecular weight.

But matching mass is necessary, not sufficient, for identity. Mass spectrometry confirms a molecular weight — it does not, by itself, confirm an amino acid sequence. Two different sequences built from the same amino acids can share an identical mass (sequence isomers), and a single point substitution can leave the total mass barely changed. This is precisely why MS and HPLC are read together: a sequence variant will usually elute at a different retention time on HPLC even when its mass looks right, and true sequence confirmation requires tandem MS (MS/MS), which fragments the molecule and reads the pieces. So when a COA shows a clean expected-versus-observed mass match alongside a single sharp HPLC peak, you have reasonable identity evidence. A mass number alone, with no chromatogram and no fragmentation data, confirms far less than it appears to.

The number nobody prints: net peptide content

Here is the gap that separates a sophisticated reader from a casual one. HPLC purity and MS identity can both look excellent while the vial contains noticeably less peptide than the label implies — because of salt, water, and counterions that the headline tests don't weigh.

Most synthetic peptides come out of reversed-phase purification as a salt, and the default counterion is trifluoroacetate (TFA) carried over from the HPLC solvents. TFA pairs with the positively charged groups on a peptide — the N-terminus and basic residues like lysine, histidine, and arginine — and every trifluoroacetate ion adds roughly 114 Da of mass that contributes nothing to the peptide itself. The 2025 Pharmaceuticals consensus paper illustrates the scale: for angiotensin, the TFA salt form can carry on the order of a quarter of its weight as trifluoroacetate. A basic-residue-rich peptide can therefore post a 99% HPLC purity and still have a meaningfully lower net peptide content once you subtract counterion, residual water, and salt. 'Net peptide content' (sometimes 'peptide content' by amino acid analysis or nitrogen determination) is the assay that actually answers 'how many milligrams of peptide are in here.' It is also the line most commonly missing from gray-market COAs. Its absence doesn't prove anything is wrong — but its presence is a strong signal that the manufacturer understands what they're selling.

Endotoxin and sterility: two different questions, often confused

Endotoxin testing and sterility testing sound interchangeable and are not. They answer different questions, and a vial can pass one while failing the other.

Endotoxin testing looks for lipopolysaccharide (LPS) — fragments shed from the cell walls of gram-negative bacteria that act as pyrogens, triggering fever and inflammatory cascades. The standard methods are described in USP chapter <85>, the Bacterial Endotoxins Test, which is built around Limulus Amebocyte Lysate (LAL), an extract historically derived from horseshoe crab blood; the newer USP chapter <86> (approved 2024, published 2025) provides a dedicated framework for non-animal recombinant reagents such as recombinant Factor C, using the same endotoxin limits and units as <85>. Endotoxin results are quantitative, expressed in endotoxin units per milliliter (EU/mL), and the crucial fact is that endotoxins are heat-stable: as the FDA's own technical guidance and depyrogenation literature note, ordinary autoclaving kills bacteria but does not destroy the toxin, which requires extreme dry heat (on the order of 250°C) or harsh chemical treatment to inactivate.

Sterility testing asks a different question — are there viable, replicating microorganisms? — and is typically qualitative: growth or no growth. The two diverge precisely because of endotoxin's heat stability: a product can be sterilized to 'no growth' yet still carry endotoxin from the dead bacteria that were killed during processing. For any material a researcher might handle in a sensitive assay, both metrics matter, and a COA that reports only one — or neither — is silent on a real risk, not reassuring about it.

Batch matching: the step that makes or breaks the whole document

A COA is only meaningful for the specific lot it tested. Everything above collapses if the certificate does not correspond to the vial in front of you. So the single most important verification step is also the most boring one: confirm that the batch or lot number printed on the COA exactly matches the batch or lot number on the physical label. Check the date as well — a certificate dated long after manufacture, or with no date, can't tell you about degradation since synthesis.

Watch for the tells of a recycled or generic certificate. A 'COA' with no batch number, or with the same batch number reused across different products, is a template, not a test record. A certificate hosted only as an image with no traceable lab name, no analyst or signatory, and no instrument or method details cannot be independently checked. The gold standard is a certificate issued by a named, ideally independent third-party laboratory, carrying the lab's identity, the method, the actual chromatogram and MS trace, and a batch number that you can match — and, ideally, that the seller will let you verify with the testing lab directly. If a vendor cannot or will not produce a batch-matched certificate with raw data, treat the absence as the answer.

A practical red-flag checklist

Pull a peptide COA up next to this list. The more boxes that go unchecked, the less the document is worth.

Missing chromatogram: a purity percentage with no HPLC trace is an unverifiable claim. Real labs show the peaks. Missing or mismatched batch number: the certificate must carry a lot number that matches your vial; a generic or absent one means the COA isn't about your material. Mass with no method context: an 'observed mass' with no theoretical mass, no instrument, and no MS/MS for sequence-sensitive products is thin identity evidence. Purity-only reporting: a high HPLC number with no net peptide content quietly ignores counterion and water weight; the vial may hold less peptide than labeled. No counterion disclosure: for basic-residue-rich peptides especially, undisclosed TFA load can mean significant dead weight. Endotoxin/sterility conflation or silence: if the product is described as suitable for sensitive work but reports neither EU/mL endotoxin nor a sterility result, the safety-relevant data simply isn't there. No traceable lab: no lab name, no signatory, no contact path to verify means nothing on the page can be confirmed. Suspiciously perfect, identical PDFs across products: certificates that look pixel-for-pixel identical except for the product name are templates.

None of these checks turn an unapproved research chemical into something approved, tested for human use, or safe — they don't. What they do is separate a manufacturer who can document their chemistry from one who is hoping you won't ask.