How to read a peptide COA: purity, mass spec, and what's missing

A certificate of analysis is the only document standing between you and an unlabeled white powder. By the end of this article you will know what each section of a COA actually measures, why "99% pure" and "5 mg" are two different claims verified by two different instruments, which risks a standard COA is silent about entirely, and the specific red flags that make a COA worthless as evidence.

Who issued it matters more than what it says

A COA is a report of laboratory tests run on a sample from a specific production lot. There are two kinds, and the difference between them is the whole game.

A vendor-supplied COA comes with the product — generated by the manufacturer, the reseller, or a lab the vendor hired. The vendor chose the sample, chose when to test it, and chose whether to publish the result. You have no way to verify that the vial in your hand came from the lot on the paper, or that the paper describes a test that ever ran.

An independent third-party test starts from the other end: a buyer sends a retail vial — one that traveled the same channel as everyone else's — to an analytical lab with no relationship to the seller. The chain of custody begins at the product, not at the factory's best sample. A directory of labs that accept these submissions is maintained on this site.

This is not a claim that vendor COAs are routinely fabricated. Many are honest. The problem is structural: a document you cannot verify does not reduce your risk, however accurate it happens to be. Treat a vendor COA as a claim. Treat an independent result as evidence.

HPLC purity: what the percentage actually measures

The headline number on nearly every COA is an HPLC purity percentage. It is worth understanding exactly what that number is.

High-performance liquid chromatography pushes the dissolved sample through a separation column; components exit at different times and pass a UV detector, typically reading at 214 or 220 nm where peptide bonds absorb. The output is a chromatogram — peaks on a timeline. "Purity 99.1%" means: of the total UV-absorbing peak area that eluted, 99.1% sat under the main peak.

Read that definition again and notice what it does not say:

• It is a relative measure, not an absolute one. It compares the main peak to the other peaks. It says nothing about how much material is in the vial.
• It only counts what the detector can see. Inorganic salts, residual solvents, water, and bacterial endotoxin either don't elute or don't absorb at the detection wavelength. A vial can be heavily contaminated with things HPLC is blind to and still read 99%.
• Co-eluting impurities hide inside the main peak. A related sequence that exits the column at the same time is counted as product.
• Purity is not identity. The wrong peptide — a cheaper compound, a failed sequence — can produce a beautiful single peak. A clean chromatogram of the wrong molecule is still the wrong molecule.

Purity is not quantity: net peptide content vs gross fill

The second claim on the label — "5 mg" — is not verified by a purity percentage at all, and this is the most common confusion in community COA readings.

Lyophilized peptide powder is never 100% peptide by weight. After solid-phase synthesis and HPLC purification, the peptide exists as a salt — most commonly a trifluoroacetate (TFA) salt, with a TFA counterion associated with each basic residue. Add bound residual water and buffer salts, and a meaningful fraction of the powder's gross mass is not peptide.

Net peptide content is the separate measurement that answers "how much actual peptide is here" — determined by amino acid analysis, nitrogen determination, or quantitative HPLC against a reference standard. Pharmacopeial monographs treat content ("assay") and purity as distinct release tests, and that separation is the tell: no professional quality system conflates them.

The practical consequence: a vial can honestly test 99% pure while containing substantially less peptide than its labeled fill weight, because purity describes the composition of the peptide fraction and says nothing about its mass. Your reconstitution math is only as good as the milligram figure you feed it — if that figure is gross fill rather than net content, every draw is off by the same factor.

A COA that states net peptide content alongside purity is a meaningfully better document than one that states purity alone.

Mass spectrometry: the right molecule, not just a clean peak

If HPLC answers "how clean," mass spectrometry answers "is it the right molecule at all."

A mass spectrometer (typically ESI or MALDI-TOF on peptide COAs) measures the molecular mass of the sample and reports it against the theoretical mass of the claimed compound. A vial sold as semaglutide has one correct answer; an observed mass matching the theoretical value is strong evidence the vial contains that molecule and not a substitute. (Semaglutide the pharmaceutical is FDA-approved as Ozempic and Wegovy; a vial of "research use only" semaglutide is not that product and carries none of its manufacturing assurances.) FDA's own guidance on synthetic peptide drug products leans on mass spectrometric methods for exactly this kind of identity and impurity characterization.

Mass spec has limits worth knowing. It confirms mass, not sequence — two peptides with identical composition in a different order weigh the same. Truncated and deletion sequences do differ in mass and appear as satellite peaks when the resolution is adequate. Full MS/MS sequence confirmation essentially never appears on a gray-market COA. But an observed mass matching theory, combined with sensible HPLC retention behavior, adds up to a reasonable identity case — which is why a COA showing purity with no mass spectrum at all has skipped the more important half of the analysis.

What a standard COA does not show

A typical COA is an HPLC trace plus a mass spectrum. Here is what that leaves out — and for an injected product, the omissions are precisely the acute failure modes.

Endotoxin deserves emphasis: it is a byproduct of bacterial contamination that survives filtration and can trigger pyrogenic reactions when injected, and it is invisible to both HPLC and MS. Sterility is likewise simply not assessed. A compound like BPC-157 — not approved for human use and sold purely as a research chemical; FDA cited safety concerns when it reviewed the substance for compounding eligibility in 2023 — reaches buyers exclusively through channels where no sterility requirement exists at any step. A pristine chromatogram is compatible with a contaminated vial.

Red flags that void a COA

Any one of these should collapse your confidence in the document:

• Recycled chromatograms. The same trace, timestamps, or integration table appearing across multiple lots — or multiple vendors. Compare against the vendor's other published COAs.
• No lot number, or a lot number that doesn't match the vial in your hand. A COA without a lot number describes nothing.
• No method information. A credible report states the column, gradient, detection wavelength, and injection date. "Purity: 99%" floating alone is an assertion, not an analysis.
• Purity with no identity data. A percentage and no mass spectrum means the more important question was never asked.
• No net peptide content anywhere on a document that claims a fill mass.
• Resolution too low to read. If you cannot make out the axes, integration values, or the fine print, that is not an accident of scanning.
• Implausible uniformity. Every lot at 99.x%, forever, with no variation, is a pattern real manufacturing does not produce.
• An unverifiable lab. No lab name, no signature, or a "lab" that is the vendor itself under another letterhead.

Using a COA well

The working posture that follows from all of this: match the lot number on the paper to the lot number on the vial before the document means anything; weight purity claims by who paid for the test; treat net peptide content as a separate question from purity; and treat sterility and endotoxin as unanswered unless a test explicitly says otherwise. Independent testing keyed to lot numbers — coordinated through the labs in the testing directory, with costs often shared across a community buying from the same lot — converts a vendor's claim into checkable evidence. Keeping the lot number recorded against each vial (a tracking app like PeptideLab stores it alongside the vial's record) is what lets a later test result, yours or someone else's, attach to the material you actually have.

A COA cannot make material safe. It can only make specific claims checkable — and knowing which claims it makes, and which it never touches, is the literacy.

Sources

• U.S. Food and Drug Administration, Guidance for Industry: "ANDAs for Certain Highly Purified Synthetic Peptide Drug Products That Refer to Listed Drugs of rDNA Origin" (2021)
• United States Pharmacopeia, General Chapter <85>: Bacterial Endotoxins Test
• United States Pharmacopeia, General Chapter <71>: Sterility Tests
• United States Pharmacopeia, General Chapters <232> and <233>: Elemental Impurities
• United States Pharmacopeia, General Chapter <467>: Residual Solvents
• U.S. Food and Drug Administration, categorization of bulk drug substances nominated for compounding under section 503A — BPC-157 review (2023)