How to Read a Peptide Certificate of Analysis (COA)

A Certificate of Analysis (COA) is a document that reports the results of laboratory testing performed on one specific batch of material. In regulated pharmaceutical manufacturing, every batch of an active ingredient, excipient, or finished product is expected to have a COA before it is released, and the document typically records identity, purity, potency, and — for injectable materials — contamination limits such as sterility and endotoxin.

The single most important idea to hold onto is that a COA describes a batch, not a product line and not a brand. The results apply to the exact lot that was sampled and tested. A COA from a batch made last year tells you very little about the vial in front of you today. This is why a genuine COA is always tied to a lot or batch number, a manufacture or test date, and ideally the name of the laboratory that ran the analysis.

It is also worth being clear about what a COA is not. It is not a guarantee of safety, it is not medical clearance, and it is not proof that a compound does anything in the body. It is a snapshot of measured chemical and microbiological attributes of one production lot. Reading it well means understanding what each measurement can and cannot tell you.

This article is educational and is written to help you interpret documents you may encounter. It is not medical advice, and nothing here should be read as guidance on using, dosing, preparing, or sourcing any substance.

Before looking at any purity figure, read the top of the document. The header is where a COA proves it is real and relevant, and it is where most worthless certificates fall apart.

Look for these identifying fields:

• Product or compound name, and often a sequence or molecular formula.
• A batch or lot number. This should match the number printed on the physical container.
• A manufacture date and/or a test date.
• The name and, ideally, the accreditation of the testing laboratory.
• A method reference or instrument identifier for each test performed.

Batch matching is the step people skip and the one that matters most. A COA is only evidence for the material whose lot number it carries. If the certificate says lot A and the vial says lot B, the document is describing something else entirely, however impressive its numbers look. A certificate with no batch number at all cannot be tied to any physical material and is effectively an unsourced claim.

Traceability is the broader principle: a trustworthy record lets you follow a result back to a dated, named test on an identified lot, run by an identifiable lab using a stated method. When any link in that chain is missing — no date, no lab, no method, no lot — the certificate loses its evidentiary value, even if the individual numbers are plausible.

High-performance liquid chromatography (HPLC), most often in its reversed-phase form, is the workhorse method for assessing how pure a peptide sample is. Reversed-phase separations with UV or mass-spectrometric detection are considered the most important analytical techniques for intact peptides and their related impurities. The instrument pushes the dissolved sample through a column that separates molecules by how strongly they interact with the column material, so that different components emerge — elute — at different times.

The output is a chromatogram: a plot with time on the horizontal axis and detector signal on the vertical axis. Each component that comes off the column appears as a peak. The target peptide is usually the large, dominant peak; smaller peaks are impurities such as deletion sequences, truncated chains, oxidized forms, or leftover synthesis by-products.

Purity is calculated by area, not height. The software integrates the area under every peak, and the purity percentage is the area of the main peak divided by the total area of all peaks — a method called area normalization. A stated purity of 98% means the main peak accounts for 98% of the total detected peak area, with everything else summing to 2%.

When you look at an actual chromatogram, check a few things. The main peak should be tall, symmetric, and clearly dominant. The baseline should be flat and quiet, not drifting or noisy. Small impurity peaks are normal and expected; what matters is their combined area. A useful sanity check is whether the reported number is consistent with the picture: if a certificate claims 99% but the trace shows several sizeable secondary peaks, the image and the number disagree, and the image is the harder thing to fake convincingly.

For research-grade peptides, a purity specification at or above about 98% is a commonly cited benchmark, though the appropriate threshold depends on the intended use and on regulatory context for therapeutic materials.

HPLC tells you how much of your sample is a single component. It does not, on its own, tell you what that component is. That is the job of mass spectrometry (MS), which measures the molecular mass of the material and lets you check it against the theoretical mass of the intended peptide.

Most peptide MS uses electrospray ionization (ESI). A crucial feature of ESI is that it produces multiply charged ions: large molecules pick up several protons, so a single compound can appear as a family of peaks at different mass-to-charge (m/z) values, each adjacent peak differing by one charge. This is why a spectrum can look confusing at first glance, and why a peak at roughly half the expected mass is often not a different molecule but the doubly charged form of the same one. Interpreting a spectrum means deconvoluting these charge states back to a single molecular weight, work that modern instrument software does automatically.

When you read the MS section of a COA, compare the reported observed mass with the theoretical (expected) mass for the sequence. They should agree closely. Small deviations are normal; a large gap can signal a wrong sequence, a chemical modification, or an adduct such as bound sodium. If the observed and expected masses are both printed and they match, that is meaningful evidence of identity.

The two methods are complementary, and the honest way to read a COA is to insist on both. A sample can be highly pure by HPLC and still be the wrong molecule. Conversely, it can show a textbook-correct mass yet hide impurities that happen to co-elute under a single HPLC peak. Purity without identity, or identity without purity, is only half of the picture.

This is the distinction that even careful readers miss, and it is where a technically true COA can still mislead.

HPLC purity is a relative measure. It answers: of the peptide-related material the detector sees, what fraction is the target? It deliberately looks past things the UV detector does not register well, such as water, salts, and the counter-ions (often acetate or trifluoroacetate) that come along from synthesis and purification.

Net peptide content is an absolute measure. It answers: of the total mass of powder in the vial, what fraction is actually peptide, as opposed to water, salt, and counter-ions? This is typically determined separately, for example by amino acid analysis or nitrogen content.

The two numbers can diverge sharply. A material can be 99% pure by HPLC yet only around 80% net peptide content, because the remaining mass of the powder is bound water and salt rather than a different peptide. Both statements are true at the same time, and neither is dishonest — they simply measure different things.

Why it matters: purity tells you about quality of synthesis and freedom from related impurities, while net peptide content tells you how much peptide you are actually holding by weight. A COA that reports a high purity but is silent on net content has told you the sample is clean, not how concentrated it is. Reading both, when both are present, gives the complete quantitative picture.

Purity and identity describe what the molecule is. A separate group of tests describes whether the material carries contaminants that matter for safety, particularly for anything intended to be parenteral. These tests appear on more thorough COAs and are worth understanding individually because they are not interchangeable.

Bacterial endotoxin testing detects lipopolysaccharide, a component of the outer membrane of Gram-negative bacteria that can cause a fever (pyrogenic) response. The standard method is the Limulus Amebocyte Lysate (LAL) assay, historically derived from horseshoe crab blood, which reacts with endotoxin and is codified in the pharmacopeial Bacterial Endotoxins Test. Results are reported in endotoxin units, and for parenteral products the acceptable limit is calculated from dose using a defined formula rather than being a single universal number.

Sterility testing asks a different question: are viable microorganisms present? The key subtlety is that sterility and endotoxin are independent. A sample can be perfectly sterile — no living organisms — and still contain endotoxin left behind by bacteria that were already killed. That is precisely why both tests exist, and why seeing only one does not cover what the other addresses.

Heavy-metal testing, usually by ICP-MS (inductively coupled plasma mass spectrometry), screens for elemental contaminants such as lead, arsenic, cadmium, and mercury that can be introduced by reagents or equipment. It is a distinct analysis that a purity chromatogram cannot capture, and it is one that lower-effort certificates often omit.

Other attributes you may see include water content (Karl Fischer titration), residual solvents from synthesis, and appearance. None of these replaces the others. A comprehensive safety picture comes from the panel as a whole, and a COA that lists only one safety test has characterized only one kind of risk.

Because a COA carries so much trust, it is a natural target for forgery and for lazy imitation. Most bad certificates fail in recognizable ways. Treat the following as warning signs rather than proof, but treat several together as a strong signal.

• No batch or lot number, or a number that does not match the physical container. Without this, the document is untethered from any material.
• No named testing laboratory, or a lab that cannot be identified or contacted. Anonymous results cannot be verified.
• No methods listed. A real report states how each value was measured (for example, RP-HPLC, ESI-MS, LAL). Bare numbers with no methods are assertions, not measurements.
• A missing, low-resolution, or cropped chromatogram. The numeric purity is trivial to type; a genuine, legible chromatogram that matches the stated number is much harder to fabricate. Be especially wary when the image is absent entirely.
• Numbers that disagree with the picture. If the trace shows prominent extra peaks but the certificate claims very high purity, believe the trace.
• Impossibly clean or suspiciously round results across every parameter, with no date. Real analytical data has texture.
• Reused documents. The same chromatogram image or the same lot number appearing across different products or dates suggests a template rather than a per-batch test.

A more subtle issue is a technically genuine COA that has been repurposed: a real report from an old or different batch, presented as if it describes current material. This is why batch matching and dates are not bureaucratic details but the core of verification. A beautiful COA for the wrong lot is worth no more than a fabricated one.

Reading a COA well ends with confirmation from outside the document itself. The certificate is a claim; verification is what turns it into evidence.

The strongest step is to check the result at the source. Many independent testing laboratories issue reports under a unique identifier and will confirm, on request or through a verification portal, that a given report number and its results genuinely came from them. Contacting the named lab and confirming that the report exists and matches defeats most forgeries, because a faked certificate typically names a lab that never ran it.

Cross-check internal consistency. The observed mass should match the theoretical mass for the stated sequence. The purity number should be consistent with the chromatogram. The batch number, dates, and product name should agree across every page. Any internal contradiction is meaningful, because a real report is generated by instruments and rarely disagrees with itself.

Prefer independent, third-party testing over in-house-only results. A manufacturer testing its own product is not disqualifying, but a certificate from an accredited external laboratory removes an obvious conflict of interest, and reports from a recognized lab carry a name and a verification path you can actually use.

Finally, keep the limits of the exercise in view. Even a fully verified, genuine COA describes the tested sample of one batch under defined methods on a given date. It cannot promise that a different vial, a later batch, or degraded material shares those properties, and it says nothing about whether a substance is appropriate or safe for any particular use. Verification tells you the document is honest about the batch it covers. What to do with that information is a separate question, and for anything health-related it is one for a qualified professional, not a certificate.