Only 30% of Vendors Publish a Lab-Verifiable COA. We Checked Our Own List.

What happened

PeptideTrust, an independent transparency scorecard, published a rubric-based rating of 20 active research-peptide vendors in July 2026, and Klow Peptide compiled the headline figures on its statistics page this month. 80% of the 20 vendors publish a Certificate of Analysis somewhere on their site, but only 30%, six of the twenty, publish one that can actually be verified at the testing lab itself, rather than just a PDF the vendor controls. Zero of the twenty publish a complete recall or complaint policy.

The scorecard grades vendors against ten weighted criteria, from whether a Certificate of Analysis exists at all up to whether it's tied to a specific batch number, names an independent lab, and can be authenticated on that lab's own site. That last one, verifiable at the source, is the strictest and rarest of the ten.

What the 30% figure is actually measuring

That gap, between a document existing and a document being independently checkable, is the actual finding here. It's a meaningfully narrower question than "does this vendor show you a lab report," and it's the reason a vendor can score well on paperwork and still land in the 70% that miss this specific criterion.

What we found when we checked our own list

We track more vendors than the scorecard's twenty, and every vendor we publish has already passed an initial review before it makes our list. So we ran our own spot-check this month, a random sample of roughly a fifth of each vendor's catalog among the compounds our readers actually shop for, 38 vendors and 121 product-COA pairs in total, checking whether a real, batch-specific document with a purity result actually exists and is reachable without a login wall.

Every one of the 38 vendors we checked had at least one product with a genuine, batch-numbered document behind it. Across the full sample, 112 of 121 product-COA pairs, about 93%, came back verified. Nine came back as real gaps, an in-stock, purchasable product with no published documentation at all, and we recorded those rather than swapping them out for something easier.

That's a much higher rate than the scorecard's 30%, and the reason isn't that the scorecard is wrong. It's that we tested a different, looser bar, whether a real document exists, on a pre-filtered list of vendors that already cleared a review before we'd link to them. The scorecard tested a stricter bar, whether that document can be authenticated at the lab itself, across a broader set of vendors that includes ones nobody has vetted. Both numbers can be true at once, and the distance between them is really a distance between two different questions.

What this means for you right now

If you're comparing vendors today, ask two separate questions, because they have two separate answers. Does this specific product have a batch-numbered Certificate of Analysis you can actually open, not a marketing line that says "third-party tested"? That's the more common bar, and most vendors we list clear it. And separately, can that document be checked against the lab's own records, not just the vendor's? That's the rarer bar, and it's worth asking a vendor directly if their product page doesn't say.

Our price comparison tool and vendor directory are where we track pricing and documentation across the vendors we monitor, so you can compare on more than sticker price alone.

What to watch next

Zero of twenty vendors in the scorecard publish a complete recall or complaint policy, which is arguably the bigger gap than the COA question and one we haven't spot-checked ourselves yet. If that number moves, or if lab-side verification becomes something more vendors build into their own product pages instead of leaving to a third party, we'll cover it.