Reading a Certificate of Analysis: What a COA States and What It Proves
A certificate of analysis is the only document that makes differences between two outwardly identical batches visible at all. Anyone able to read one can tell within a few minutes whether a supplier knows its goods or merely encloses an attractive PDF. The advice to look for a complete certificate appears in many texts — how to recognize a good one, rarely. This article works through a certificate from top to bottom: which entries carry weight, which technical terms appear in it, which shortcomings are common, and where its informative value ends.
What a certificate of analysis is and who issues it
A certificate of analysis, usually abbreviated to COA, is a test report. In it, a laboratory documents which sample it received, by which method it examined that sample, and which measured values came out of the examination. That is precisely where its strength lies: it is a record of measurements, not an evaluation and not a release — and for that reason it can be checked.
It is issued by the examining laboratory, meaning either an independent contract laboratory or an in-house one. The difference is considerable, because a test report from the laboratory of the same company that places the goods on the market stands in a different relationship to them than one from a third party. Which of the two applies is evident from the document — provided the laboratory is named at all.
The header data: the most load-bearing part of the document
The upper section of a test report looks like administrative trimming and in truth carries everything else. A complete certificate names there:
- the unambiguous designation of the sample and of the product,
- the batch or lot number to which the testing refers,
- the date the sample was received and the date of testing,
- the name and address of the laboratory,
- a sequential report number,
- information on who drew the sample and how it reached the laboratory.
From this follows the single most important check of all: a certificate of analysis without a batch reference belongs to no goods. Something was indeed measured, but it can no longer be said what. A test report is not a seal attached to a product name; it is a record of one sample from one particular lot. Conversely: anyone who can match a batch number on the packaging with the same number on the certificate holds a seamless connection between document and goods.
The same applies to the dates. The date of receipt says when the material arrived at the laboratory, the date of testing when the measurement was taken. If either is missing, it remains open what condition the sample was in at the time of measurement.
Accreditation: what ISO/IEC 17025 means
Many test reports carry a reference to the standard DIN EN ISO/IEC 17025. It lays down general requirements for the competence, the impartiality and the consistent operation of laboratories — from personnel and methods through measurement uncertainty to reporting of results and data integrity. In Germany, the Deutsche Akkreditierungsstelle (DAkkS) accredits to this standard as the national accreditation body.
One subtlety deserves a second look: what is accredited is not a laboratory as a whole, but a defined scope. Which methods and which test objects are covered is set out in the annex to the accreditation certificate. A laboratory can therefore be accredited without the specific examination falling within that scope. A test report without accreditation is not for that reason wrong — the measurement may be perfectly sound. It is simply harder to verify from the outside. An accredited report with a matching scope is the stronger piece of evidence, and that is the real reason the question is worth asking.
The alkaloid section: a number without a reference basis is meaningless
The substantive core of a certificate for plant material is the list of the constituents determined — for Mitragyna speciosa typically mitragynine and 7-hydroxymitragynine, sometimes further alkaloids. What is decisive here is not the number but its reference basis. A percentage figure can refer to:
| Reference basis | What the percentage figure then states |
|---|---|
| Dry weight | Proportion of the dried material as a whole |
| Fresh weight | Proportion of the material including residual moisture |
| Total alkaloids | Proportion within the alkaloid fraction, not of the material |
| Weight by weight, weight by volume | The usual form for liquids and solutions |
The same quantity of substance yields entirely different figures depending on the reference. A proportion of total alkaloids is regularly orders of magnitude higher than the same proportion of dry weight, because the alkaloids make up only a small part of the material. If the reference basis is missing, a percentage figure can neither be placed in context nor compared.
That this is no academic quibble becomes clear from a look at official limits: they always name their reference basis expressly. The US agencies, for instance, phrase their 2026 proposed threshold for 7-hydroxymitragynine on a dry weight basis and place a second, per-article criterion alongside it for liquids and portioned forms. Knowing the reference basis of a certificate is therefore what makes it possible to compare a figure with a rule in the first place.
With liquids and portioned forms there is the additional question of whether a figure refers to a volume or to a unit — a concentration and an amount per unit are not the same thing.
The method: what HPLC and LC-MS/MS mean
A complete certificate names the method applied, and does so precisely enough for it to be identified. Two abbreviations turn up particularly often.
HPLC stands for high-performance liquid chromatography. The sample mixture is forced through a separation column in a stream of solvent; the components travel through it at different speeds and arrive at the detector separated in time. Identification is based on the time a substance needs for that passage, compared with a known reference substance.
LC-MS/MS couples such a separation with two mass spectrometers arranged in series. There the molecules are recognized by their mass and by their fragments. The method is more demanding, but it reaches lower measuring ranges and distinguishes similar molecules more reliably — which carries particular weight with closely related alkaloids.
Then there is calibration. A content figure arises through comparison with a reference standard of known purity; if that standard is named in the report, the number can be traced back to its foundation. Why the method has to be stated follows directly from this: two figures from two laboratories are comparable only if it is known how they came about.
Limit of detection and limit of quantification
This is the technically most important part of the article, and it changes how many reports should be read. Test reports frequently carry wordings such as "not detectable", "n.d." or "< LOQ". These do not mean that the substance sought is absent. They mean that with this method it could no longer be captured below a certain threshold.
The German standard DIN 32645 distinguishes three quantities for this purpose:
- The limit of detection is a purely qualitative statement: below it, it can no longer be decided whether anything is present at all.
- The limit of identification lies above it and serves, under the standard, as a statement about a maximum possible content.
- The limit of quantification is the smallest concentration that can still be stated numerically with a defined accuracy. Only from that point on is a quantitative statement possible.
Two things follow from this. First, a finding of "not detectable" without the corresponding limit is incomplete, because it remains unclear how deeply anyone looked in the first place. Second, findings from different laboratories are comparable only if their limits are comparable — a more sensitive method finds traces that a less sensitive one reports as not detectable. The difference then lies in the method, not in the goods. A certificate that discloses its limits is for that reason the more informative one, even if at first glance it looks less polished.
Measurement uncertainty: the number behind the number
Every measurement result has a spread. Measurement uncertainty states the range within which the true value lies with a defined probability; determining and reporting it is among the requirements ISO/IEC 17025 places on laboratories. It becomes practically relevant wherever a result sits close to a line: a value just below or just above a limit cannot be judged conclusively without its uncertainty. A report that states it permits exactly that assessment — one that omits it suggests a sharpness the measurement does not possess.
The contaminants section
Besides the constituents, a comprehensive test report covers three further groups.
Heavy metals. The usual ones are lead, cadmium, mercury and arsenic. In the EU food sector the reference values are the maximum levels of Regulation (EU) 2023/915 on contaminants in food, which replaced the earlier Regulation (EC) No 1881/2006. For products that are not foodstuffs it does not apply directly; a test report should then ideally name the frame of reference against which the values are placed.
Microbiology. What is typically examined is the total viable count, yeasts and molds, and individual organisms such as Salmonella or Escherichia coli. Mycotoxins, aflatoxins in particular, also fall into this area.
Pesticide residues. The framework in the EU is Regulation (EC) No 396/2005, which harmonizes maximum residue levels for pesticides in products of plant and animal origin. If no specific maximum level has been set for a combination of active substance and product, a default value of 0.01 milligram per kilogram applies. For a test report what counts here is not only the result but also how many active substances were screened for in the first place — a screening panel covering several hundred active substances says considerably more than one covering a dozen.
The most common shortcomings
| Shortcoming | Why it devalues the document |
|---|---|
| No batch reference | The report cannot be assigned to any specific goods |
| No testing or receipt date | The measurement has no point in time |
| Percentage figures without a reference basis | The number can neither be placed in context nor compared |
| No method named | The origin of the number remains open |
| "Not detectable" without a stated limit | Unclear how deeply the measurement went |
| A cropped image instead of a complete document | Header and footer data including the signature are missing |
| Laboratory name without an address | The issuer cannot be identified |
What a good certificate makes visible
The benefit can be summed up in one sentence: a complete test report makes quality differences verifiable that a product otherwise does not show. Two powders of the same color can differ markedly in their alkaloid profile, in heavy metal content and in microbial count — that stands in the certificate and nowhere else. Equally visible is how thoroughly the testing was done: a report with a named method, stated limits of quantification, declared measurement uncertainty and a broad pesticide panel comes from a different standard of quality than a single line with a round percentage figure. Anyone who knows what to look for can make that distinction — and that is the practical yield of this article.
Where its informative value ends
A test report says nothing about safety and nothing about effects. Nor does it make any statement about whether goods may lawfully be placed on the market — that is a legal question, not an analytical one.
It says exactly one thing: what was measured in this one sample on this one day with this one method. Every statement beyond that comes not from the document but from whoever presents it. Anyone reading a certificate therefore reads the header first, then the reference bases, then the limits — and the numbers last.
Further sources: the information from the Deutsche Akkreditierungsstelle on testing and calibration laboratories under DIN EN ISO/IEC 17025 (dakks.de), Regulation (EU) 2023/915 on maximum levels for certain contaminants in food (EUR-Lex), Regulation (EC) No 396/2005 on maximum residue levels of pesticides (EUR-Lex), and the DEA notice of July 6, 2026, in which the proposed threshold is phrased on a dry weight basis (Federal Register).
Further reading
- Heavy Metals & Pesticides in Kratom: Why Quality Testing Matters
- Kratom and its alkaloids
- What is mitragynine?
- What is 7-hydroxymitragynine?
- Kratom varieties explained simply
Legal notice
This article is for informational purposes only and does not constitute legal advice. Its content is not intended to encourage consumption. Laws may change; the applicable regulations and information from official bodies are authoritative. Image source: https://www.kratoein.com/