How to Read a Peptide COA: Line by Line

A peptide COA (certificate of analysis) is a lab report for one production lot. The header says who tested what, and when. The results table reports identity (mass spec) and purity (HPLC). Some also report net peptide content, water, counter-ion, endotoxin and heavy metals. Every line is only as good as the lot number and the lab behind it.

The table below is a reading template, not a real certificate. Use it as a checklist when you open a vendor's PDF. If a COA fills in only the identity and purity rows, you know exactly what was not tested. Knowing which rows are missing is half the skill.

If you only check three things, check these: the lot number matches your vial, the lab is a named third party, and the purity number comes with a chromatogram. Our peptide quality guide covers vendor-level green and red flags. This page stays on the document itself.

HPLC purity is the share of the detector signal that belongs to the main peak. The sample runs through a column, its components separate, and a UV detector draws a peak for each one. Peptide methods commonly read the signal near 220 nm. Purity (area %) is the main-peak area divided by the total area of all integrated peaks.

Two published methods show the typical setup: USP's peptide reference-standard work monitored HPLC at 220 nm (McCarthy et al., 2023), and a melanotan II study used UV at 218 nm (Breindahl et al., 2015). Area % is a relative number. It compares peaks with each other. It does not weigh anything, so it says nothing about how much water or salt is in the powder. That is why USP's mass-balance approach measures peptide-related impurities, counter-ion, water, residual solvents and non-combustible residue as separate items (McCarthy et al., 2023).

Yes. A single smooth peak only shows that nothing separated out under that one set of conditions. Some impurities elute at almost the same time as the main peptide and have the same mass. One example is a version with one amino acid flipped to its mirror-image form during synthesis. Peak-purity checks look for anything hiding under the main peak.

Racemization during solid-phase synthesis can create such diastereomeric impurities (D'Hondt et al., 2014). Isomers with the same mass-to-charge ratio are not readily told apart by mass spectrometry, so they have to be separated chromatographically. A 2023 study built a two-dimensional LC-MS strategy specifically to assess main-peak purity of pharmaceutical peptides (Petersson et al., 2023). A simpler, older approach uses a diode-array detector to compare absorbance ratios at several wavelengths across the peak. A co-eluting species changes the ratios, though the method is relatively non-specific (Castledine et al., 1991).

On its own, very little you can check. "HPLC tested" or "99% pure" on a label or product listing says someone ran a test at some point. It doesn't say which lot, which lab, which method, or what the result was. The evidence is a lot-matched COA with a chromatogram, from a lab you can contact.

Published test purchases show how far a label can drift from the lab result. In a 2024 study of semaglutide sold online without a prescription, the vials' measured purity was 7.7–14.37% against the 99% claimed on the labels (Ashraf et al., 2024). In a study of melanotan II vials bought from three online shops, each sold as 10 mg, the vials contained 4.32–8.84 mg. Vials from two of the shops held 4.1–5.9% unknown impurities (Breindahl et al., 2015).

Purity tells you how much of the peptide material is the right peptide. Net peptide content tells you how much of the powder's weight is peptide at all. Lyophilized peptide powder also holds water, counter-ions such as acetate or TFA, and residual solvents. So a lot can be 99% pure by HPLC and still be well under 100% peptide by weight.

USP assigns purity to its bulk peptide reference materials by mass balance. Every detectable non-target component is measured and subtracted from 100%: peptide-related impurities, counter-ion, water, residual solvents and non-combustible residue (McCarthy et al., 2023). Water is measured by Karl Fischer titration (USP chapter <921>). Labs can also quantify peptide content directly: by HPLC assay against a reference standard, by quantitative NMR (qNMR), or by amino acid analysis. A USP multi-laboratory study compared all three on oxytocin (Li et al., 2019).

Most peptides are sold as salts. Synthesis and purification use trifluoroacetic acid, so many peptides come out as trifluoroacetate (TFA) salts. That stays true unless the maker swaps the counter-ion for acetate or chloride. The salt-form and counter-ion lines tell you which salt you have and how much of the weight it takes up.

Because of cleavage and purification conditions, solid-phase-synthesized peptides are mainly obtained with TFA present. Cationic peptides come out as TFA salts, and exchanging that counter-ion for one such as acetate is often required. Reverse-phase HPLC and ion-exchange resin are among the documented ways to do it (Roux et al., 2008). Approved peptide drugs carry a range of counter-ions, and the choice affects physicochemical properties and formulation (Sikora et al., 2020). Residual trifluoroacetate is also on the standard list of peptide-related impurities (D'Hondt et al., 2014).

Why labs care: in a cell-culture study, trifluoroacetate (TFA) at 10^-8 to 10^-7 M inhibited proliferation of osteoblasts and chondrocytes. The authors recommended converting peptides to a hydrochloride or equivalent salt before testing their biological effects (Cornish et al., 1999). For reading a COA, the practical point is simpler. A "TFA salt" lot and an "acetate" lot of the same peptide are different materials. Their weights are not interchangeable.

Find two numbers: the theoretical mass for the peptide's sequence and the observed mass the instrument measured. They should agree within the tolerance the lab states. Electrospray instruments often detect multiply charged ions, so a raw m/z value can be about half or a third of the molecular weight. A good report shows that conversion.

Example from the literature: melanotan II was identified from its doubly charged [M+2H]²⁺ ion at m/z 513, along with retention time and five fragment ions (Breindahl et al., 2015). Doubling 513 and subtracting the two added protons gives a molecular mass of about 1,024 Da. USP's reference-standard testing used LC-MS/MS both to confirm the peptide's mass and to verify its amino-acid sequence (McCarthy et al., 2023).

Most peptides are made by solid-phase peptide synthesis (SPPS), which builds the chain one amino acid at a time. Each step can leave a characteristic by-product. A missing amino acid gives a "deletion" sequence and an extra one gives an "insertion." Others include mirror-image (D) residues, leftover protecting groups, oxidation, and dimers. These appear as the small peaks around the main HPLC peak.

A 2014 review in the Journal of Pharmaceutical and Biomedical Analysis catalogued the impurity types most often found in peptide medicines (D'Hondt et al., 2014):

A stronger COA lists impurity peaks individually, by retention time and %. It sometimes names them by mass, for example a "des-" (deletion) variant. Note that ICH's general impurity guideline for new drug substances, Q3A(R2), explicitly does not cover peptides (ICH Q3A(R2), scope). So don't expect a peptide COA to cite Q3A thresholds for its impurity peaks.

Endotoxin is lipopolysaccharide from the outer membrane of Gram-negative bacteria. It can stay in a product after the bacteria are gone, so a "sterile" result and a low endotoxin result are separate tests. A useful endotoxin line gives a number in EU/mg (endotoxin units per milligram), the test method, and the limit applied.

The methods. USP chapter <85>, "Bacterial Endotoxins Test," describes the gel-clot, photometric and kinetic test methods. FDA has accepted the Limulus amebocyte lysate (LAL) test for more than 30 years (FDA, Pyrogen and Endotoxins Testing: Q&A). A 2020 proficiency study described LAL assays as the gold standard. Recombinant factor C (rFC) tests, which drop the horseshoe-crab blood components, performed comparably to LAL in a six-year proficiency program (Piehler et al., 2020).

The limit. There is no single universal pass number. Endotoxin limits are calculated as K/M, where K comes from the threshold pyrogenic human dose of 5 EU per kg and M is the maximum dose per kilogram (Malyala & Singh, 2008). A limit therefore depends on the dose and route the lab assumed. That is why "Endotoxin: Pass" with no number, method or limit can't be checked.

Element by element, with a number for each, measured by an ICP technique (ICP-MS or ICP-OES). Look first for the four ICH Q3D "Class 1" elements: arsenic, cadmium, mercury and lead. A single generic "heavy metals" line is a sign the lab used the old, non-specific test that current pharmacopeial chapters replaced.

USP <232>/<233>, EP 2.4.20/5.20 and ICH Q3D ended the non-specific metals analysis of the old USP <231> and EP 2.4.8. They require substance-specific quantitative analysis with lower detection limits, which in practice means ICP-AES or ICP-MS (Fischer et al., 2014). ICH Q3D classifies As, Cd, Hg and Pb as Class 1: human toxicants with little or no use in pharmaceutical manufacturing, which must be evaluated in every risk assessment (ICH Q3D(R2)). Q3D's permitted daily exposures (PDEs) for these four elements:

PDEs are daily totals for a finished drug product. They are not a per-vial pass mark. When reading a COA, the useful question is whether each element was measured and reported with a number and a method. If a COA has no elemental line at all, that doesn't make the product contaminated. It means the question is unanswered.

Compare the vial against the COA's appearance line, typically a short description such as "white lyophilized powder." The ideal freeze-dried cake is uniform. But a less-than-ideal look (shrinkage, cracks, a cake pulled away from the glass) can be purely cosmetic. A vial that looks wet, sticky, discolored, or nothing like the description is worth a question to the vendor.

A cross-industry commentary on freeze-dried products noted that cake appearance may or may not matter for quality. A non-ideal cake is sometimes just an inherent result of the formulation, the vial and the freeze-drying process. The authors proposed a harmonized nomenclature for cake defects and a risk-based way to set acceptance criteria (Patel et al., 2017). In other words, appearance is a screening check, not a purity test. A good-looking cake can still fail HPLC, and an odd-looking one can pass.

Storage after delivery is its own topic: see our peptide storage guide.

The biggest red flags are a lot number that doesn't match your vial, a lab you can't identify or contact, and a purity number with no chromatogram behind it. After those come vague "pass" results with no numbers, recycled report images, and missing salt forms. Any one is a reason to ask questions. Several together are a reason to buy elsewhere.

If the paperwork can't settle it, test the product itself. Our PeptideVerify page lists independent labs that accept samples, and the peptide quality testing guide explains which tests to order.

Short definitions for the terms that show up on peptide certificates and lab reports.

COA (certificate of analysis)

A lab report of test results for one specific lot.

Lot / batch

One production run. A COA applies to one lot only.

HPLC

High-performance liquid chromatography: separates a sample's components so each can be measured.

Chromatogram

The HPLC trace: detector response plotted against time, one peak per separated component.

Retention time

When a peak comes off the column. The main peak's time is method-specific.

Area %

A peak's share of the total integrated peak area. This is how HPLC purity is reported.

Peak purity

A check that the main peak contains only one component (diode-array spectra or a second method).

Mass spectrometry (MS)

Measures molecular mass. Used on a COA to confirm identity.

m/z

Mass-to-charge ratio. A doubly charged ion shows up at roughly half the molecular mass.

Monoisotopic vs average mass

Mass from the most common isotopes vs the natural isotope mix. Check which one the lab used.

Net peptide content

The percent of the powder's weight that is peptide, excluding water, counter-ions and solvents.

Counter-ion / salt form

The ion paired with a charged peptide, e.g. trifluoroacetate (TFA), acetate or chloride.

Karl Fischer (KF)

The titration used to measure water content.

SPPS

Solid-phase peptide synthesis, the usual way peptides are made.

Deletion / insertion sequence

A synthesis impurity missing an amino acid, or carrying an extra one.

Diastereomer

A version of the peptide with one residue in its mirror-image (D) form. Same mass, different shape.

Lyophilized cake

The freeze-dried solid in the vial.

Endotoxin / EU

Bacterial lipopolysaccharide, measured in endotoxin units (EU), usually per mg.

LAL / rFC

Limulus amebocyte lysate and recombinant factor C: the two main endotoxin test types.

ICP-MS / ICP-OES

Element-specific methods for measuring metals such as As, Cd, Hg and Pb.

PDE

Permitted daily exposure: ICH Q3D's daily limit for an elemental impurity in a drug product.