The Research Peptide COA Standards Report is an independent framework for judging whether a peptide certificate of analysis (COA) actually means anything. Most COAs in this market look identical at a glance and mean wildly different things in practice. We built a 10-point COA Trust Scorecard so researchers — and the vendors who serve them — can score any certificate on the same scale instead of taking “99% pure” on faith.
This report is free to cite and reference. If you publish in the research-compound, longevity, or biohacking space, you’re welcome to link to or quote the Scorecard below.
Why a COA standard is needed
A certificate of analysis is supposed to be evidence. In research peptides it is too often decoration. The same PDF gets reused across unrelated batches for 6–12 months; purity figures appear with no test method named; “third-party tested” links to a file the vendor hosts on its own domain with nothing to verify it against. When we audit a new batch, the first thing we look at is not the purity number — it’s whether the document can be independently checked at all. Most can’t.
The result is a market where a 99.7% on one COA is rigorous and a 99.7% on another is meaningless, and a buyer has no way to tell them apart. The Scorecard exists to close that gap.
The COA Trust Scorecard (10 points)
Score one point for each criterion a certificate meets. We consider 8/10 the practical floor for research-grade sourcing; anything under 5 should be treated as unverified marketing.
| # | Criterion | What a real COA shows | Common failure |
|---|---|---|---|
| 1 | Independent lab | Tested by a named third-party lab (e.g. Janoshik Analytical), not the vendor | “In-house tested” or no lab named |
| 2 | Independently verifiable | A code you can check on the lab’s own system, not the vendor’s site | PDF hosted only on the vendor domain |
| 3 | Batch-specific | A batch/lot number that matches the vial you receive | One COA reused across many batches |
| 4 | Recent test date | Tested within a sensible window of the batch you’re sent | A COA dated 8–12 months earlier |
| 5 | Method named | States the method — HPLC for purity, MS for identity | “Purity: 99%” with no method |
| 6 | Identity confirmed | Mass spec (LC-MS) confirming the molecule is what’s claimed | Purity only, no identity test |
| 7 | Quantified purity | A specific figure (e.g. 99.6%), not “high purity” | Vague descriptors, no number |
| 8 | Net peptide content | Distinguishes peptide content from total mass (salt/water) | Headline purity that hides low net content |
| 9 | Endotoxin / sterility layer | An LAL or sterility result beyond HPLC purity | HPLC only — the layer most vendors skip |
| 10 | Legible & complete | Full, unredacted, readable document — not a cropped image | Screenshot of a corner, key fields cut |
What purity numbers actually mean
The single most misread field on a COA is the purity percentage, because the interesting range is almost invisibly narrow. In our experience the difference between a 98.2% and a 99.7% HPLC result is not cosmetic — in sensitive assays that ~1.5% of unidentified material is exactly the variable that hurts reproducibility. Treat the bands like this:
| HPLC purity | How to read it |
|---|---|
| ≥ 99% | Research-grade target. What a serious supplier aims for batch to batch. |
| 98–99% | Usable but worth questioning — ask what the remaining 1–2% is. |
| 95–98% | Below research-grade for most work; impurity profile matters a lot here. |
| < 95% or “high purity” | Unquantified or low — treat as unverified. |
Purity is not identity. A high HPLC figure tells you the sample is clean; it does not tell you it’s the right molecule. Only mass spectrometry (LC-MS) confirms identity. A COA with purity but no identity test is half a document — which is why criteria 5 and 6 are scored separately above.
Match-batch vs pooled testing
The criterion that separates most vendors is #3. There are two models in this market:
- Pooled / reused COA: one test result is presented for many production runs, sometimes for most of a year. The document you’re shown may have nothing to do with the vial in your hand.
- Match-batch COA: a separate independent test per production batch, with a batch number that matches your label. The certificate describes your material.
We consider single-COA, pooled sourcing the single biggest red flag in this category — not because the number is always wrong, but because it’s unfalsifiable. You can read more in our cornerstone guide, Janoshik Match-Batch: What It Means + How to Verify.
How the report was built (methodology & data note)
The Scorecard is a qualitative framework, not a claim of proprietary lab capacity. Bastion is a sourcing operation: our compounds are produced by suppliers who submit each batch to Janoshik Analytical (Prague) for independent HPLC purity and identity testing, and every resulting COA is verifiable on Janoshik’s own public system. Across the supplier batches we have verified to date (n = 6), independent HPLC purity has ranged 99.49%–99.82% (mean 99.65%). That is a small, honestly-disclosed sample used here only to illustrate the bands above — not an industry-wide dataset. The Scorecard’s value is the method, which applies to any vendor’s certificate, including ones we have nothing to do with.
You can inspect the underlying certificates on our lab results page, and apply the Scorecard yourself to any vendor — start with our research compounds or a competitor’s.
Cite this report
Researchers and writers are free to reference the COA Trust Scorecard. Suggested citation: “COA Trust Scorecard, Bastion Peptides Research Peptide COA Standards Report (2026).” Want to partner on a data project or earn through referencing our verification work? See our affiliate program.
Research Use Only. Not for human consumption. All compounds referenced are intended exclusively for in-vitro laboratory research.