A Certificate of Analysis is the one document in this market that contains verifiable information rather than adjectives. It is also the document most buyers glance at, see numbers and a laboratory letterhead, and accept as proof. This guide is about reading it properly — what each section is, which tests matter, and the ways a COA can be technically real and still tell you nothing about the product you are about to buy.
Safety & legal note. Kratom is not approved by the FDA for any use, and nothing here is medical advice. This article explains laboratory documentation, not how anyone should use kratom. Kratom is banned in some US states, is not for anyone under 21, and can cause dependence with heavy use. Check your local law before ordering.
What a COA actually is
A COA is a laboratory's report on one specific batch of one specific product. A lab receives a sample, runs a defined set of tests, and states what it found.
That definition contains the two limits people miss. A COA covers a sample, not the whole batch — and it covers that batch, not the product line. It is evidence about a lot of powder, not a certificate of virtue attached to a company.
Who issues it, and why third-party matters
The report should come from an analytical laboratory independent of the vendor. The credential to look for is accreditation to ISO/IEC 17025, the international standard for testing-laboratory competence — an accredited lab has had its methods and quality systems assessed by an outside body.
Three sources you will encounter, in descending order of usefulness:
| Source | What it's worth |
|---|---|
| Independent accredited third-party lab | The standard. Named lab, its own letterhead and report format. |
| Vendor's in-house lab | Useful for process control, but the party making the claim is grading it. |
| Unattributed "lab results" graphic | Not a COA. A design element. |
A genuine COA names the laboratory, and you can look that laboratory up. If the document does not identify who ran the tests, there is nothing to verify.
The tests that matter
Not every COA runs every panel, and the gaps are where risk hides. A COA showing only alkaloid content is a potency report, not a safety report.
Alkaloid content
Usually reported as a percentage by weight, and typically covering:
- Mitragynine — the most abundant alkaloid in the leaf.
- 7-hydroxymitragynine — present in far smaller amounts in raw leaf, and the figure that matters most for spotting adulteration. Independent testing has found products spiked with synthetic 7-hydroxymitragynine, so an unusually high number on something sold as plain leaf is what to notice.
This section tells you whether the label is honest about what is in the bag. It matters more on extracts, because a multiplier like "50x" is a marketing figure with no standard definition behind it, while an alkaloid percentage is a measurement.
Heavy metals
The panel to look for covers the standard four: lead, arsenic, cadmium, and mercury. These accumulate in the plant from soil and water rather than being added — a sourcing issue, and one the FDA has issued warnings about in kratom products.
We are not publishing threshold numbers here: applicable limits depend on product category, jurisdiction, and assumed serving size, and quoting a figure we cannot properly source would be worse than quoting none. What you can do without any thresholds is confirm all four metals were tested, that each line carries both a result and a stated limit, and that the report marks each as pass or within spec. A COA that lists a limit alongside every result is doing the interpretation for you — that is the format you want.
Microbials
Kratom is a dried agricultural product, and salmonella has driven kratom recalls in the US. A microbial panel typically covers total aerobic count, yeast and mold, E. coli, and Salmonella — the last normally reported as absent/detected rather than as a number.
Moisture
Low-glamour, genuinely useful. Excess moisture in stored plant material is what allows mold to grow, so the result is an indirect indicator of how the product was dried and stored.
Pesticides and residual solvents
Less commonly run on plain leaf. Solvent testing becomes relevant for extracts, where solvents may be used in processing. An extract COA with no solvent panel is a real gap.
Batch and lot numbers: the field that makes it meaningful
The lot number is what connects the document to the product. Without it, a COA is a report on an unidentified quantity of powder somewhere in the vendor's history.
Find the lot number on the packaging or product page, then confirm the COA carries the same one. If the vendor publishes a searchable COA library, look it up yourself rather than accepting an emailed PDF. If the number on the bag appears nowhere in the vendor's published results, you have a COA for something you are not buying.
Dates: three of them, and they are different
| Date | What it tells you |
|---|---|
| Sample received | When the lab got the material |
| Test / report date | When the analysis was run and issued |
| Batch or pack date | When the product itself was produced |
The relationship between them is the useful part. A report date long after the sample was received, or a batch date far older than the report, both warrant a question. A COA with no date at all is unusable — you cannot tell whether it describes current inventory or something from years ago, and undated is the most common form of COA theater precisely because it never expires.
There is no universal shelf life for a COA. The practical standard: it should correspond to the batch being sold right now.
Reading one, in order
- Find the lab name and confirm it is not the vendor.
- Match the lot number to the product you are buying. If it doesn't match, stop here.
- Check the dates — is the report plausibly about current inventory?
- Check which panels ran — alkaloids only, or contamination too?
- Read the results and limits columns together. A number with no limit beside it is not interpretable.
- Look for pass/in-spec markings and footnotes. Footnotes are where deviations get disclosed.
- Confirm the product description matches what you are buying — strain, format, powder versus extract.
Step 7 catches a common substitution: a COA for plain leaf presented on the page for an extract, or a COA for a different strain entirely.
Red flags
- No lot or batch number — the report cannot be tied to your product.
- No date — a report that never expires never proves anything current.
- Vendor-issued rather than third-party — the seller grading its own work.
- No laboratory named — nothing to verify.
- Product description doesn't match what you are buying.
- Alkaloids only, no contamination panels — potency dressed up as safety.
- Results with no limits column — numbers you cannot interpret.
- A cropped screenshot instead of the full multi-page report.
- One COA covering an entire product line — batches vary; that is why batch testing exists.
- "COA available on request" followed by no COA on request.
What a COA cannot tell you
Over-trusting the document is its own failure mode:
- It describes a sample, and sampling is imperfect.
- It cannot prove the powder in your bag came from the lot on the report — that link depends on the vendor's traceability, not the lab's work.
- It says nothing about anything the panel did not test for.
- It says nothing about whether kratom is appropriate for anyone. It is a chemistry document.
A COA is necessary but not sufficient, which is why GMP qualification carries more weight in our scoring than published COAs alone: an audited process is what makes any individual batch report trustworthy.
The bottom line
A useful COA names an independent accredited lab, carries a lot number matching your product, is dated recently enough to describe current inventory, and covers contamination as well as potency. Anything missing one of those four is not proof — it is a document that resembles proof.
Our guide to spotting clean kratom covers the signals around the document, and our vendor rankings weight lot-matched third-party testing above all else.