Researched and fact-checked in-house against primary literature and regulator records. Not reviewed by a named clinician — how we work.
Evidence-rated reference Updated August 2026
We sell nothing. No vendor sponsorship. Editorial policy
pepteyes .com

Product Quality & Sourcing

How to Read a Peptide Certificate of Analysis

A certificate of analysis reports identity and purity for one sample on one day. Here is what those numbers measure, what the document leaves out, and what it can never establish.

What a certificate of analysis is, and what it is not

A certificate of analysis is a report on a specific batch. It lists a set of tests, the procedure used for each, an acceptance criterion, and the result obtained on a sample drawn from that batch. In the vocabulary the ICH Q6A guideline established, a specification is the combination of tests, procedures and acceptance criteria; the certificate is the record that one batch was measured against that specification and passed. It is a factual document about a sample at a point in time.

It is not a licence, a marketing authorisation, a safety evaluation, or a warranty that anything is suitable for use in a person. It carries no pharmacology, no toxicology, and no regulatory status. In the research-peptide market it is often the only technical artefact anyone produces, so it ends up shouldering the evidentiary weight of an approval it was never designed to carry.

The productive way to read one is line by line, asking four questions of each entry. What physical property was measured? By what method? On which sample? And what would have failed to show up if it were wrong? Answered honestly, that turns most certificates into a short list of things measured and a much longer list of things not measured. The absences are where the reading actually happens.

The header block: lot, dates, and who ran the tests

Before any analytical result comes an identification block: product name, stated sequence, batch or lot number, quantity, dates of manufacture and analysis, sometimes a retest date, and the issuing laboratory. This block is checkable against itself without any instrument at all. Does the stated sequence correspond to the named compound? Does the molecular formula match the sequence, and the theoretical mass match the formula? Internal contradictions here are informative, because they are errors a genuine analysis would not have produced.

Dates matter more than they look. Analysis long after manufacture describes a batch that has already aged, and a retest date is a stability claim only as good as the programme behind it, which a research certificate rarely has.

The issuing party is the most consequential line and the least standardised. A certificate from the manufacturer reports the manufacturer's own testing. One on independent laboratory letterhead means a different party ran the assay, which closes exactly one gap. It does not mean an independent party selected the sample, collected it, or observed what happened to the batch afterwards. Third-party testing, as the phrase is normally used, is independent analysis of a submitted specimen, not independent sampling.

Identity: what mass spectrometry confirms and what it cannot

The identity line on almost every peptide certificate is a mass spectrometry result: electrospray ionisation produces multiply charged ions that are deconvoluted to a single mass and compared against the theoretical mass calculated from the stated sequence. Close agreement is a useful screen, ruling out a wholly different compound, a grossly truncated chain, or an inert filler substituted for the peptide.

It does not confirm sequence. Intact mass is order-blind, so any rearrangement of the same residues gives an identical mass and a scrambled chain passes. Leucine and isoleucine are exactly isobaric and can never be separated by mass. Glutamine and lysine differ by roughly 0.036 daltons, which a high-resolution instrument resolves and a nominal-mass instrument does not. Deamidation shifts the mass by about one dalton and methionine oxidation by sixteen, so the tolerance window a certificate accepts decides which of these it can flag at all.

Establishing the actual order of residues takes tandem mass spectrometry with fragment ion assignment, or classical sequencing chemistry, and those results are uncommon on research certificates. The point sharpens for modified peptides: semaglutide carries an aminoisobutyric acid substitution and a fatty diacid chain attached through a linker, and a correct intact mass is compatible with that chain sitting at the wrong position. Mass answers what, at best, not where or in what order.

Purity: how to read an HPLC area percent

The purity figure is nearly always reversed-phase high-performance liquid chromatography with ultraviolet detection, usually near 214 nanometres, where the peptide bond itself absorbs. The instrument integrates every peak, and the reported purity is the main peak area divided by the total integrated area. The phrase for this is area percent, and the two words do a great deal of work.

Area percent assumes every impurity responds to the detector in proportion to the main peak, which is never exactly true. A contaminant with few chromophores is undercounted; one rich in tryptophan or tyrosine is overcounted. More importantly, the denominator contains only material that both absorbed and eluted. Inorganic salt, residual water and many solvents pass through effectively invisible or leave in the void volume, and a species that never comes off the column under the gradient used does not exist as far as the number is concerned.

So a certificate reading 99.4 percent means that 99.4 percent of the ultraviolet-absorbing material eluting under this gradient, on this column, at this wavelength, appeared in one peak. Without the chromatogram, column chemistry, gradient, wavelength and run time, the figure is not checkable by anyone, and a bare percentage with no trace attached is a claim rather than a result. The difference between a shallow gradient over forty minutes and a fast generic one over eight is the difference between resolving closely related impurities and averaging them into the main peak.

The impurities a single chromatographic run will not resolve

Solid-phase synthesis fails in characteristic ways, and its characteristic failures are the hardest impurities to see. A deletion sequence, missing one internal residue because a coupling step did not go to completion, differs from the target by a single amino acid and often elutes very close to it. A truncated sequence, where chain growth stopped early, may be further away or may not be. Whether either appears as a separate peak is a property of the method, not of the material.

Stereochemical impurities are worse. Racemisation during activation and coupling, most readily at cysteine and histidine, produces diastereomers with the same formula, the same mass, and retention behaviour that can be nearly indistinguishable on a standard reversed-phase column. Detecting them requires hydrolysing the peptide and analysing the liberated amino acids by a chiral method, a separate assay research certificates essentially never include. Aggregates are a third category, invisible under reversed-phase conditions that dissociate them and detectable instead by size exclusion chromatography.

The general principle is orthogonality. Two methods that separate on different physical grounds tell you far more than one method reported to an extra decimal place. A certificate showing a single reversed-phase run and a single mass result has established two things about the material, not an impurity profile.

Purity is not how much peptide is in the vial

Purity is a relative measure: what fraction of the peptide-like material present is the intended species. Net peptide content is absolute: what fraction of the powder in the container, by weight, is peptide at all. These answer different questions, and a certificate can report the first while saying nothing about the second.

Net peptide content comes from an orthogonal quantitative method, most commonly amino acid analysis, in which the peptide is hydrolysed and the released residues are quantified against standards. Elemental nitrogen determination and quantitative nuclear magnetic resonance serve the same purpose. All give a weight-per-weight figure that no chromatographic area percent can supply or substitute for.

The consequence is arithmetic rather than subtle. Lyophilised synthetic peptides, especially basic sequences purified as trifluoroacetate salts, commonly fall well short of being pure peptide by mass once counterion and water are counted. A container labelled with a nominal milligram quantity of material reported at high chromatographic purity can hold appreciably less peptide than the label suggests, and the purity figure will not have moved. Most research certificates omit net peptide content, so the number most relevant to how much active substance is present is the one missing.

Water and counterion: the mass that is not peptide

Residual water in a lyophilised cake is measured by Karl Fischer titration, the compendial method for water in solids. Freeze-dried peptides retain some moisture, and hygroscopic material takes up more once a container is opened. Water is invisible at 214 nanometres and contributes nothing to an area percent, but it contributes directly to the weight on the label.

Counterion is the larger term. Reversed-phase purification normally runs in trifluoroacetic acid, so the peptide leaves that process as a trifluoroacetate salt with counterions associated with each basic site. Trifluoroacetate has a formula weight near 114, so on a peptide with several basic residues the counterion contribution is not a rounding error. Quantifying it takes ion chromatography or fluorine nuclear magnetic resonance, and converting to another salt form is a deliberate extra step. A certificate reporting purity but not salt form leaves the composition of the powder partly undefined.

Salt form is not a technicality to regulators either. When compounded semaglutide products proliferated, the FDA stated plainly that semaglutide sodium and semaglutide acetate are different active ingredients from the semaglutide base its approvals cover. That is the regulatory expression of the same chemical point: a document naming a salt describes a different substance from one naming the free base, and the two are not interchangeable because the peptide chain matches.

Sterility, endotoxin and particulates: absent by design

The tests that determine whether something can be injected are almost never on a synthesis certificate, and that is not an oversight. Compendial sterility testing, the bacterial endotoxins test using limulus amebocyte lysate, and particulate limits for injections are attributes of a finished sterile product, set by the manufacturing environment, the filling process and the container closure. They are not properties of a synthesis, and no identity or purity result implies anything about them.

Endotoxin deserves separate emphasis because the intuition about it is usually wrong. It is a heat-stable bacterial cell wall component that survives autoclaving and passes through sterilising-grade filters. Material can therefore be sterile, in the sense of containing no viable organisms, and still be pyrogenic. It can equally be 99 percent pure by chromatography and carry a substantial endotoxin burden, because purity and endotoxin load are unrelated quantities measured by unrelated assays.

Lyophilisation compounds the misunderstanding: freeze-drying removes water, but it does not sterilise and does not destroy endotoxin. A certificate reporting identity, purity and water content has said nothing about microbiological quality, and the honest conclusion from that silence is that the question was not asked.

The question the document cannot answer: does it describe this vial?

Every result on a certificate applies to a sample that reached a laboratory. The step from there to any particular container in the world is not an analytical step, and no instrumentation inside the report can supply it. The only bridge is a batch number on the paper matching one on a label, and in the research market both are printed by the same party.

The document is also trivially forgeable. A certificate is a PDF; text in a PDF is editable, letterhead is copyable, and a genuine certificate for one good batch can be reissued alongside every later batch without anything on its face changing. Some laboratories publish verification portals precisely because the document cannot authenticate itself, and one naming no verifiable issuer offers no route to checking.

The counterfeit semaglutide episodes made the stakes concrete. The FDA seized counterfeit Ozempic that had entered the legitimate supply chain in late 2023, and the World Health Organization issued a medical product alert on falsified semaglutide identified across multiple regions in 2024. Some counterfeit units contained insulin rather than semaglutide. In each case the falsified product travelled with convincing packaging and paperwork, which is the point: documentation is the easiest part of a product to fake, because faking it requires no chemistry. A certificate is evidence about a sample, and its connection to a specific vial is an assertion the analytical chemistry on the page is silent about.

What a regulated release package looks like by comparison

For an approved product, batch release under good manufacturing practice generates a file that a single certificate merely summarises: identity by orthogonal methods, a quantitative assay, related substances measured against thresholds qualified during development, water content, sterility, endotoxin, particulates, container closure integrity, and stability data supporting the expiry. Behind that sit a quality system, validated methods, retained samples, and a regulator empowered to inspect the facility and the records. The certificate is meaningful because of the system it emerged from.

The FDA's guidance on abbreviated applications for certain highly purified synthetic peptides referencing recombinant listed drugs shows the characterisation regulators actually expect. The question it poses is not whether a single purity number clears a threshold, but whether the complete impurity profile of the synthetic material is comparable to that of the reference product, with impurities identified and their relevance argued. That is a materially different exercise from reporting one area percent.

For compounds with no approval anywhere, the comparison has no other side. BPC-157 holds no marketing authorisation in any major jurisdiction; the FDA placed it in the category of its 503A bulk drug substances evaluation reserved for substances raising significant safety concerns, and the World Anti-Doping Agency has named it on the Prohibited List since 2022. TB-500, a synthetic fragment marketed in reference to thymosin beta-4, is likewise unapproved and falls in the prohibited growth factor class in sport. For these there is no compendial monograph, no established specification and no qualified impurity threshold, so a certificate is written against acceptance criteria the issuer selected. That does not make the measurements false; it means passing them shows consistency with a self-chosen standard, and reading the page as equivalent to a regulated release document mistakes the form of the evidence for its weight.

What we still don't know

Every claim above has a limit. These are the questions the current evidence does not answer.

  • No published survey establishes what proportion of certificates circulating in the research-peptide market correspond to an actual analysis of the batch they accompany, rather than a reused, edited or fabricated document.
  • It is unmeasured how often such material would reproduce its stated purity if retested by an independent laboratory using an orthogonal separation instead of a repeat of the supplier's own gradient.
  • For unapproved peptides there is no qualification threshold for impurities, so the level at which a deletion sequence, truncated chain or diastereomer becomes biologically consequential in humans has never been determined.
  • Whether the diastereomer content typical of non-pharmaceutical peptide synthesis alters receptor activity or immunogenicity in humans has not been studied, partly because that content is rarely quantified in the first place.

Common questions

Does a certificate of analysis mean a peptide is safe to inject?
No. A certificate reports chemical attributes of a sample, typically identity by mass and purity by chromatography. Safety for injection depends on sterility, endotoxin load, particulate content and the manufacturing environment, none of which a synthesis certificate ordinarily tests, and on pharmacology and toxicology, which no certificate addresses at all.
What does 99 percent purity on a peptide certificate actually mean?
It usually means an area percent from one reversed-phase HPLC run with ultraviolet detection near 214 nanometres: the main peak accounted for 99 percent of total integrated peak area. Only material that both absorbed at that wavelength and eluted under that gradient is in the denominator, so salt and water are excluded, co-eluting impurities merge into the main peak, and the figure says nothing about how much of the powder by weight is peptide.
Can mass spectrometry prove the peptide has the correct sequence?
Not on its own. Intact mass confirms the molecule weighs what the stated sequence predicts, which rules out a wholly different compound or gross substitution. It cannot detect a rearranged residue order, cannot distinguish leucine from isoleucine at all, and separates glutamine from lysine only on a high-resolution instrument. Establishing sequence order requires tandem mass spectrometry with fragment assignment or classical sequencing chemistry.
Why do certificates rarely report sterility or endotoxin results?
Because those are attributes of a finished sterile product rather than of a chemical synthesis, generated by a filling operation and its environment rather than by a purification step, so a supplier not manufacturing under sterile conditions has nothing to report. The absence states what was tested, not that the result was favourable, and neither high purity nor lyophilisation implies microbiological quality, since endotoxin survives filtration and heat.
Is third-party testing the same as independent verification?
No. Third-party testing means an independent laboratory performed the assay on a specimen it received. It does not mean an independent party chose which material to submit, collected the sample, or confirmed that the batch tested is the batch subsequently distributed. It closes the gap of who ran the instrument and leaves open what was submitted and what happened afterwards.
How can anyone tell whether a certificate is genuine?
From the document alone, largely not. Internal consistency can be checked without a laboratory: whether the stated sequence matches the named compound, whether formula and theoretical mass agree with that sequence, whether batch and dates are coherent, and whether a chromatogram with its method conditions is attached rather than just a number. Failing those checks is informative; passing them is not proof, because a competent forgery passes them too.

What this is based on

Named sources, with what each one actually showed. We link live literature searches rather than a frozen citation list, so you can check the current record yourself.

  1. ICH Q6A specifications test procedures and acceptance criteria for new drug substances and new drug products — Defines a specification as the combination of tests, analytical procedures and acceptance criteria a batch must meet, which is the framework a certificate of analysis reports against. find on PubMed
  2. FDA guidance ANDAs for certain highly purified synthetic peptide drug products that refer to listed drugs of rDNA origin — Sets out the impurity characterisation and profile comparison regulators expect for synthetic peptides, an exercise well beyond reporting a single chromatographic purity figure. find on PubMed
  3. USP General Chapter 71 Sterility Tests — Establishes the compendial sterility test for injectable products, an attribute independent of chemical identity and purity results. find on PubMed
  4. USP General Chapter 85 Bacterial Endotoxins Test — Defines the limulus amebocyte lysate based limit test for pyrogenic endotoxin, which is unrelated to chromatographic purity and survives sterilising filtration. find on PubMed
  5. USP General Chapter 788 Particulate Matter in Injections — Sets limits on subvisible particulate contamination in injectable products, another finished-product attribute absent from a synthesis certificate. find on PubMed
  6. USP General Chapter 921 Water Determination — Specifies Karl Fischer titration as the compendial method for residual water, the measurement that quantifies moisture retained in a lyophilised cake. find on PubMed
  7. USP General Chapter 1052 Biotechnology-derived Articles Amino Acid Analysis — Describes amino acid analysis after hydrolysis, the orthogonal quantitative method underlying net peptide content determinations. find on PubMed
  8. ICH Q3C guideline for residual solvents — Classifies residual solvents and sets exposure limits, an attribute rarely reported on research peptide certificates despite solvent use throughout synthesis and purification. find on PubMed
  9. FDA alert on compounded semaglutide products using semaglutide sodium and semaglutide acetate salt forms — The agency stated these salt forms are different active ingredients from the semaglutide base covered by its approvals, establishing that salt form changes substance identity. find on PubMed
  10. FDA warning on counterfeit Ozempic semaglutide injection seized from the US drug supply chain in December 2023 — Counterfeit units entered legitimate distribution with convincing packaging and documentation, and some were found to contain insulin rather than semaglutide. find on PubMed
  11. WHO Medical Product Alert on falsified semaglutide 2024 — Documented falsified semaglutide batches identified across multiple regions, demonstrating that plausible accompanying paperwork does not establish product authenticity. find on PubMed
  12. FDA 503A bulk drug substances evaluation categorising BPC-157 — The agency placed BPC-157 in the category reserved for substances raising significant safety concerns, so no compendial specification or qualified impurity threshold exists for it. find on PubMed
  13. WADA Prohibited List entry for BPC-157 and growth factors including thymosin beta-4 derivatives — Establishes that BPC-157 has been explicitly named as prohibited in sport since 2022 and that thymosin beta-4 derivatives such as TB-500 fall within the prohibited growth factor class. find on PubMed

Peptides covered here

Terms used in this article

Certificate of Analysis (COA)
A certificate of analysis is a lot-specific document listing the tests run on a batch, the specification for each and the result obtained, and it is a claim to be checked rather than proof.
Approved vs Cleared vs Authorised
Approved, cleared and authorised are three different FDA outcomes resting on different evidence, and only approval means the agency reviewed data showing the product works for its stated use.
Batch and Lot Number
A batch or lot number is the unique code identifying material made in one manufacturing cycle, linking a physical vial to its production records and to any recall that later names it.
Third-Party Testing
Third-party testing is analysis of a product by a laboratory independent of the seller, and its value depends entirely on who chose the sample and which assays were actually run.
Mass Spectrometry Identity Confirmation
Mass spectrometry identity confirmation ionises a sample and compares its measured mass-to-charge ratio against the mass calculated from the sequence the product claims to be.
Amino Acid Residue
An amino acid residue is what remains of an amino acid once it has been joined into a chain and a molecule of water has been lost, and residue count is how peptide length is stated.
Dalton (Da)
The dalton is the unit of molecular mass equal to one twelfth of a carbon-12 atom and numerically identical to grams per mole, the standard scale for sizing peptides and proteins.
Non-Natural Amino Acid
A non-natural amino acid is any residue outside the twenty encoded by the genome, incorporated synthetically to block proteolysis, enforce a conformation, or add chemistry the standard set cannot provide.
High-Performance Liquid Chromatography (HPLC)
High-performance liquid chromatography separates a mixture on a packed column under pressure and reports each component as a peak, whose relative area is the basis of stated peptide purity.
Peptide Bond
A peptide bond is the amide linkage joining the carboxyl group of one amino acid to the amino group of the next, formed with the loss of water and remarkably resistant to spontaneous hydrolysis.
Solid-Phase Peptide Synthesis (SPPS)
Solid-phase peptide synthesis builds a chain one residue at a time on an insoluble resin, so excess reagents wash away by filtration instead of requiring each intermediate to be isolated.
Truncated Sequence
A truncated sequence is a synthesis impurity whose chain assembly stopped early, leaving a shorter molecule missing residues from one end rather than from the middle of the sequence.

Read next

This article is educational information, not medical advice. It cannot account for your medical history, medications, or risk factors. Do not start, stop or change any treatment based on it. Speak to a qualified healthcare professional who knows your case. We publish no dosing protocols for unapproved compounds and link to no supplier.

← All 25 articles