The document everyone asks for
Ask how to judge a peptide source and the answer arrives in three letters: check the COA. This site says it too. The availability layer only lists a compound when it comes with a lot-specific certificate of analysis. But a requirement to have the document is not the same as knowing how to read it, and the gap between those two is where most of the false comfort lives.
A certificate of analysis is a report on a sample: someone took material from one manufacturing lot, ran a set of tests, and wrote down the results. In regulated manufacturing, the list of tests and the limits each must meet are fixed in advance as the product’s specification (ICH Q6A), and the certificate is the record that a lot met it. Outside that system, a COA is whatever the seller chose to test and chose to show. Reading one well means asking three questions of every line: what did this test actually measure, what is it blind to, and does it describe the vial in front of me?
Start at the top: is this about your vial?
Before any number, read the header. A certificate should name the peptide, the lot or batch number, the date of testing, the laboratory that did the work, and the methods used. The lot number is the only thread connecting the paper to the powder: if the number on the certificate is not the number on the vial, the document is about some other material. A certificate with no lot number, or one shared across every order for months, is a brochure.
Then ask who ran the tests. A manufacturer’s own certificate and an independent laboratory’s report are different kinds of evidence, and the stronger packets include both. Look for the raw output, too: the chromatogram trace and the mass spectrum, not only a table that summarizes them. A summary can be typed; a trace with its baseline noise, retention times and integration marks is harder to invent and far easier to sanity-check.
Even a genuine, matching certificate describes the sample that was tested on the day it was tested. It cannot describe what happened to your vial afterwards: the fill, the shipping, the weeks in a warehouse.
Purity: one peak's share of the light
The headline number is almost always HPLC purity. The sample is pushed through a column that separates its components, and a detector measures how much ultraviolet light each one absorbs as it comes out, usually at a wavelength where the peptide bond itself absorbs. “99% pure” means the main peak made up 99% of the total peak area on that trace. It is an area percentage: a share of what the detector saw.
That definition has consequences. Things that do not absorb at that wavelength are invisible, which includes water, salts and the counterions left over from synthesis. And things that leave the column at the same moment as the main peak are counted as the main peak. Solid-phase synthesis reliably produces close relatives of the intended molecule: sequences missing one amino acid, sequences with an extra one, residues flipped to their mirror-image form, oxidized or still-protected side chains (D’Hondt et al., 2014). Some of those separate cleanly. Some sit under the main peak unless the method was built to pull them apart. The same powder can read 99% on a short, generic gradient and noticeably lower on a slower one designed for that peptide.
So the purity figure is real information with a narrow meaning. It tells you the synthesis and purification were not a disaster. It does not tell you how much peptide is in the vial, and it does not tell you the vial is clean in any biological sense.
Identity: the mass matches, the sequence might
The second standard line is mass spectrometry. The instrument weighs the molecule, and the certificate compares the observed mass with the one calculated from the intended sequence. A match is strong evidence you have the right compound and not a different peptide altogether, which is a failure that genuinely happens in unregulated supply.
What a single mass cannot do is tell apart molecules built from the same atoms. Two residues in swapped positions weigh the same. So does a peptide in which one amino acid has racemized to its D-form during synthesis, a known side reaction (D’Hondt et al., 2014) that can change how the molecule fits its receptor. Confirming the order of residues takes fragmentation data (MS/MS) or other methods that routine certificates rarely include. Identity by mass means “the right weight”, which is most of the way to “the right molecule” and not all of it.
Content: why a 10 mg vial is not 10 mg of peptide
Here is the line most certificates leave out, and the one that matters most for anyone doing arithmetic. Purity asks what share of the peptide-like material is the intended peptide. Content asks what share of the powder is peptide at all.
A lyophilized peptide is a salt. Its basic groups are paired with counterions, commonly acetate or the trifluoroacetate left by purification, and the dried cake holds on to some water. None of that shows on the HPLC trace. A powder can therefore be 99% pure and still be meaningfully less than 99% peptide by weight, and a vial labeled by the gross weight of powder holds less active molecule than the label suggests. Content is measured separately, by amino acid analysis or nitrogen determination, and reported as net peptide content when it is reported at all.
The counterion is not just dead weight. Trifluoroacetate at very low concentrations inhibited the proliferation of bone and cartilage cells in culture (Cornish et al., 1999), enough to distort experiments that blamed or credited the peptide for what its salt was doing. A certificate that names the salt form is telling you something useful.
Then there is the simplest content question of all: how much went into the vial. A certificate tests the bulk lot, not the fill. When Belgian regulators profiled the falsified peptide products they seized most often, they found wide variation in the amount per unit alongside low purity (Janvier et al., 2018). This is the quiet reason the reconstitution arithmetic people rely on can be precise and still wrong: the calculation starts from a number on a label that nobody verified.
| Line on the COA | What it measures | What it cannot see |
|---|---|---|
Lot / batch number Lot 2611-A | Which production run the tested sample came from. | Whether your vial came from that run. Only a matching number on the vial links the two, and a number is easy to print. |
Identity (MS) Observed mass matches theoretical | That the molecular weight is right. | Same-mass errors: transposed residues, D-amino acids from racemization. |
Purity (HPLC) 99.1% at 220 nm | The main peak's share of everything the UV detector saw. | Water, salts, counterions; impurities hiding under the main peak; anything biological. |
Peptide content Often absent | How much of the powder's weight is actually peptide. | Nothing about what else the peptide fraction contains. Read it together with purity. |
Endotoxin (LAL) Often absent | Bacterial endotoxin in the tested sample, in endotoxin units. | Live organisms, and any vial that was not sampled. |
Sterility Usually absent | Whether the sampled units grew anything. | The rest of the lot. Sterility comes from the process, not the test. |
The lines that are usually missing
HPLC and mass spectrometry are chemistry. They are blind to biology. A vial can be chemically immaculate and carry bacteria, or carry endotoxin, the fragments of bacterial cell wall that cause fever and worse when injected and that survive the death of the bacteria that made them. Endotoxin has its own test, the limulus amebocyte lysate assay, with results in endotoxin units and limits set by dose and route (FDA pyrogen and endotoxin guidance). If a certificate does not report it, it was not shown to be absent.
Sterility is harder still, because it cannot really be tested into a product. A sterility test destroys the vials it examines and samples only a handful from a lot, so it can catch gross contamination and little else. In licensed manufacturing, sterility is assured by the process: validated filtration, aseptic filling, environmental monitoring. A sterility line on a certificate is welcome, but it is a statement about a few sampled units.
The same seized-product survey found something no purity figure would reveal: arsenic and lead, with several samples carrying levels up to ten times the limit set for injectable medicines (Janvier et al., 2018; ICH Q3D). Elemental impurities and residual solvents each need their own assay. A two-line certificate has simply not looked.
What no PDF can prove
Suppose the certificate is complete: matching lot, named laboratory, traces attached, content, endotoxin and sterility all reported. It is still a document, and three things sit outside any document.
The first is the link to your vial. Nothing on the page can show that the powder you hold came from the lot that was tested. When researchers bought semaglutide from websites selling it without a prescription and had it analyzed, the products’ measured purity fell far short of the figure on their labels, the vials held more semaglutide than declared, and endotoxin was detectable (Ashraf et al., JAMA Netw Open, 2024). Paperwork and contents had come apart.
The second is time. Peptides degrade: they deamidate, oxidize and aggregate, faster with heat and moisture. A result from the day of testing does not follow the vial through a summer in transit.
The third is the system around the test. A medicine’s certificate is the last page of a controlled process that regulators inspect. Material sold for research use is not made, filled or released under that process, which is the substance of the FDA’s warnings about unapproved versions of GLP-1 drugs (FDA). A good certificate narrows the uncertainty. It does not convert one kind of product into the other, a distinction the supply chain itself makes hard to see.
A certificate of analysis is best at ruling things out. It can show that a lot was not the wrong molecule and not grossly impure. It gets weaker with every step between that tested sample and the liquid in a syringe.
Reading order
Put together, a certificate is read from the outside in: first whether it applies, then what was measured, then what was never measured.
- Match the lot. The lot number on the certificate must be the one on the vial, with a test date, a named laboratory and stated methods. No match, no information.
- Read purity as a share of the trace. HPLC purity is the main peak's percentage of UV-visible material on one method. It is blind to water, salt and counterions, and to impurities that co-elute.
- Read identity as a weight. A matching mass rules out the wrong peptide. It does not rule out a same-mass variant such as a swapped or mirror-image residue.
- Look for content. Net peptide content, the salt form and water tell you how much of the powder is peptide. Without them, the milligrams on the label are the weight of the cake.
- Check what is missing. Endotoxin, sterility, elemental impurities and residual solvents each need their own test. An absent line is an untested property.
- Remember what paper cannot do. No certificate proves your vial came from the tested lot, survived shipping unchanged, or was made under a process built for injection.
Keep going
- How to reconstitute peptides — what happens after the vial is opened
- How to read a peptide dose — four kinds of milligram
- Where the powder comes from — the supply chain behind the certificate
- Cataloged vs. reachable — the sourcing standard this site applies
- The complexity ladder — why longer peptides are harder to make cleanly
- Methodology — how this reference grades evidence and sources
- Ask the research agent about peptide quality testing
Educational reference on analytical quality testing, summarized from public scientific literature and regulatory guidance and simplified in places. Not individualized medical advice, and not an endorsement of using research-grade material in people. Specific studies and guidance documents are named to explain the science; verify any claim against the linked primary sources.