How Research Peptide Vials Fail Quality Control — And the Six Checks That Catch It

The research peptide market runs on a single number. "99% purity" appears on nearly every product page, and for most buyers it is the whole of quality assurance. It is a real measurement. It is also, on its own, close to uninformative — and the published analytical literature is now detailed enough to show precisely why, and precisely how vials fail.

What the Testing Data Actually Show

A 2026 preprint analysed 6,285 consumer-submitted third-party test results across 14 compounds including BPC-157, TB-500, retatrutide, semaglutide, tirzepatide, GHK-Cu, PT-141 and CJC-1295. Depending on which acceptance model was applied, between 41.6% and 71.1% of samples failed basic quality criteria. Median purity was high — 99.80% — but median measured abundance was 101.80% of label with an interquartile range of 95–109%. The authors’ conclusion is the important one: dosing accuracy, not gross impurity, was the dominant deficit. Pass rates varied enormously by compound: against the stricter manufactured-product model, TB-500 passed in 7.5% of cases and CJC-1295 in 10.7%.

A caveat worth stating, which most coverage does not: this is voluntarily submitted data, not a random sample, and it is a preprint rather than peer-reviewed work. The authors themselves describe their failure rates as a lower bound.

The cleanest demonstration of the purity/content gap comes from a 2024 study in JAMA Network Open, which test-purchased semaglutide from online sellers operating without prescription. Three lyophilised vials were delivered; three orders were simply non-delivery scams. LC-MS on the delivered material showed a single chromatographic peak with no peptide-like impurities. Measured polypeptide concentration was 14.37%, 8.97% and 7.70% against a labelled claim of at least 99% purity. Sterility testing found no viable organisms — and endotoxin was present in all three at 2.16–8.95 EU/mg.

A separate certificate-integrity study ordered 98 synthetic peptides from two international suppliers at a specified minimum 95% purity. All 98 arrived with certificates stating at least 95%. In-house quality control found 43 of 98 actually met it. The authors identified the mechanism: the supplier’s own chromatographic method clearly lacked selectivity. For one peptide, the supplier’s chromatogram claimed at least 95%; a selective method found 80.6%. For another, the main peak was not the ordered sequence at all — it was the intramolecularly disulfide-cyclised form, 2.1 Da lighter. A pure-looking trace of the wrong molecule.

And sometimes the compound simply is not there. Polish official medicines control laboratory analysis of 601 samples seized between 2020 and 2024 examined a 63-sample post-cycle-therapy subset in which 34.9% did not contain the declared active substance at all — 20.6% contained none whatsoever, and 14.3% contained a different, undeclared one. Within the peptide-specific portion, of 16 samples declared as human chorionic gonadotrophin, 37.5% contained no active substance — mannitol only.

What Each Analytical Method Actually Proves

The gap between what a test measures and what buyers assume it measures is where most of the confusion lives. RP-HPLC at 210–220 nm measures purity: the main peak area as a percentage of total peak area. It has four blind spots, each evidenced above.

• It is a ratio, not a mass. It cannot tell you how much peptide is in the vial.
• It only counts what absorbs UV and elutes. Water, counterion, inorganic salt and bulking agents are invisible to it.
• It is method-dependent. Same peptide, 80.6% or ≥95% depending on column and gradient.
• The main peak may not be your peptide. Disulfide-scrambled and cyclised species can present as a single clean peak.

LC-MS and high-resolution MS confirm identity — molecular weight, and with MS² the sequence through fragment ions. But they do not give quantity without a calibrated reference standard, do not distinguish D- from L-amino acid epimers (identical mass), do not reliably distinguish leucine from isoleucine, and an intact-mass check with a loose tolerance will not catch an N-terminal glycine (+57 Da) or alanine (+71 Da) extension — both identified in real seized material.

• Amino acid analysis — net peptide content. Acid hydrolysis to free amino acids, then quantification. This is the test that answers how many milligrams of peptide are actually in the vial, and its absence is the single largest information gap in grey-market documentation.
• Karl Fischer titration — water content. Lyophilised cake is hygroscopic, and water is both dead mass and a degradation accelerant.
• Ion chromatography — counterion. TFA from reversed-phase purification cannot be completely removed; the FDA expects counterion content and identity on a drug substance specification.
• LAL or recombinant Factor C — endotoxin. Endotoxin is heat-stable and is not removed by sterile filtration. A vial can be sterile and pyrogenic at once. It must be measured directly and reported as a number with a unit.

The Eight Ways a Vial Goes Wrong

Solid-phase synthesis has a characteristic failure catalogue. The question that matters commercially is whether routine RP-HPLC catches each one.

• Deletion sequences (incomplete coupling): usually caught, but short polar deletions can co-elute. 74% of all identified impurities in the 98-peptide study involved at least one deletion.
• Insertion sequences (excess amino acid driving double coupling): often caught.
• Racemisation / D-epimers: frequently NOT caught — identical mass, similar retention. Needs chiral analysis. Cysteine racemisation of ~50% has been documented under the wrong base.
• Oxidation of Met, Trp or Cys: usually caught, since +16 Da is more polar and elutes earlier.
• Deamidation of Asn/Gln via succinimide: only partially — the +0.98 Da shift sits at the edge of low-resolution MS.
• Aggregation and high-molecular-weight species: NOT caught. Largely invisible to analytical RP-HPLC; needs SEC or light obscuration.
• Disulfide scrambling: deceptively — presents as a clean single peak that is the wrong molecule.
• Residual protecting groups (tBu, Trt, Pbf adducts): usually caught, since the hydrophobicity shift is large.

Two of these — aggregation and disulfide scrambling — are the ones regulators cite most often. The FDA’s stated rationale for placing several peptides in its higher-risk compounding category is immunogenicity due to the potential for aggregation and peptide-related impurities. Aggregates are precisely what a purity chromatogram does not show you.

What a Certificate Has to Contain to Mean Anything

There is an established standard, and it is not a matter of opinion. ICH Q7 §11.4 and EU GMP Part II require a certificate of analysis to carry the name of the substance, the batch number, release and expiry dates, a list of each test performed including acceptance limits, numerical results rather than "pass" or "conforms", a dated signature by authorised personnel, and the name of the company or laboratory.

The FDA’s expectation for a peptide drug substance specification goes further: identity by a combination of orthogonal methods, a stability-indicating assay, related substances by orthogonal methods, plus water content, residual solvents, elemental impurities, counterion content and identity, bacterial endotoxins and microbial limits. Measured against that, a certificate reporting only HPLC purity and MS mass confirmation omits six of the nine expected attributes.

Impurity thresholds are what give a purity figure meaning. The European Pharmacopoeia monograph for synthetic peptides sets reporting at 0.1%, identification at 0.5% and qualification at 1.0%. The FDA’s synthetic peptide guidance requires applicants to identify each peptide-related impurity at 0.10% or greater, and justification for any new impurity must include data showing it does not contain sequences with increased MHC affinity — potential T-cell epitopes. A bare percentage with no impurity table cannot speak to any of this.

One point of honesty: there is a great deal of industry talk about fabricated and recycled certificates, and no peer-reviewed study or regulatory publication quantifies it. What can be said with evidence is narrower and more useful — a certificate lacking a lot number, numerical results or acceptance limits fails on its face against ICH Q7, and supplier-generated purity figures have been shown empirically to overstate.

The Six Checks, in Order

• Does the certificate carry a lot number, a test list with acceptance limits, numerical results, and a dated signature? Anything less is not a certificate of analysis in the regulatory sense.
• Does it report net peptide content or assay — not just purity by HPLC? Without it the ratio is known and the mass is not. This is the single most common omission.
• Does it report water content and counterion identity and quantity? These are the two largest non-peptide mass sinks in a lyophilised vial.
• Does it report endotoxin as a number with a unit? Purity predicts pyrogenic load with an R² below 0.01. If it is not measured, it is not known.
• Is identity confirmed by MS² or peptide mapping rather than intact mass alone? Glycine- and alanine-extended analogues have been found in seized material precisely because a loose mass check passes them.
• Was the analysis performed on the specific lot supplied, by a laboratory independent of the seller, with the method disclosed on the certificate? This is the check the published evidence supports most strongly — and the one most rarely satisfied.

The Underlying Point

Purity, content, identity and pyrogen load are four independent measurements. Treating one as a proxy for the others is not a small analytical imprecision; it is the specific error the 2024 test-purchase study documented in its purest form — a flawless chromatogram sitting on top of a vial that was roughly 90% something else. Read "99% purity" as a statement about a chromatogram, not about a vial. Everything else follows from that.

Related Research