Three numbers, three questions
A certificate of analysis (COA) for a research peptide usually reports several values. They answer different questions, and confusing them is the most common mistake buyers make.
Chromatographic purity (HPLC %) answers one question: what fraction of the peptide-related material is the target peptide? It is measured by high-performance liquid chromatography, which separates the main peptide from related impurities. Independent labs such as Janoshik Analytical report this as a percentage. For research-grade material, values in the high-90s are commonly cited as good.
Net peptide content answers something else. Of the powder in the vial, how much is actually peptide? The remainder is water, counter-ions (often acetate or trifluoroacetate), and salts left from synthesis. A vial can be 99% pure by HPLC yet contain noticeably less than its labeled mass in actual peptide, because purity and content are independent measurements. HPLC area-percent purity does not include the bound water and counter-ion salt present in the dried material, and content is typically determined by a separate method such as quantitative amino acid analysis.
Identity answers whether this is the right molecule at all. It is confirmed by mass spectrometry, which checks molecular weight against the expected value. Purity is meaningless if the identity is wrong.
What a grade does and doesn't tell you
Aggregated grades roll these measurements, plus dose accuracy, into a single letter. The A–E ratings published by Finnrick, which we cite on our vendor pages with attribution, are one example. Per Finnrick's published methodology, each tested sample gets a 0–10 score combining purity, quantity accuracy versus the label claim, and batch-information quality. Samples with too few tests are left unrated (X), and a separate F grade is reserved for documented fraud. A grade like that is useful for scanning, but the underlying COA is where the detail lives. It summarizes someone else's testing. It doesn't replace reading the report.
Reading the report itself
A few things worth checking, drawn from how analytical labs describe their own methods:
- Who ran the test. A COA hosted on an independent lab's own domain is more verifiable than a PDF or screenshot supplied by the seller. Janoshik, for example, lets buyers confirm a report by entering its verification key, which appears on the certificate.
- The date and batch. Testing applies to the batch tested. A COA from a different batch is evidence about a different production run.
- The method. HPLC purity and MS identity should both be present. A purity number with no identity confirmation is incomplete. The methods themselves follow established pharmacopeial conventions; the United States Pharmacopeia's general chapter on chromatography, USP <621>, was last updated effective May 2025.
We aggregate and link these reports rather than re-grade them ourselves. The point is to make the original testing easy to find and read, not to replace it. None of this is a quality judgment for human use. Independently tested or not, these are research compounds, not products approved for human consumption.