A COA, or Certificate of Analysis, records the tests performed on an identified sample and the results obtained. It may include measured concentrations, analytical methods and a sample or batch number. On its own, it does not prove effectiveness, complete safety or suitability for everyone.
The useful question is whether the document supports the conclusion you want to draw. Start with the sample and batch, then read the dates, units and testing scope. This guide explains how to connect those details without filling gaps with assumptions.
What does a Certificate of Analysis contain?
A COA brings together identification, testing and results. Look for the laboratory name, report reference, sample description and any batch or lot number. Then find the individual tests, methods, units, reported values and acceptance criteria.
The World Health Organization’s model certificate describes these elements for pharmaceutical starting materials and finished products. It provides useful context for understanding report structure, but it is not a universal legal checklist for cosmetics, supplements or CBD products. (WHO, Model certificate of analysis, 2018.)
Read the complete file rather than a cropped image of one percentage. Footnotes, explanatory codes and limitations may appear on later pages. Check whether the page numbering indicates that anything is missing before relying on the visible results.
Step 1: match the product, sample and batch number
The product name is your starting point, not the whole identification check. Compare the batch number on the packaging with the code in the report. Read every character, including suffixes, rather than accepting a close match.
Next, check what was tested. A raw ingredient, concentrated extract and finished mixture are different materials. An ingredient report does not automatically describe the composition of a later diluted product. If different codes appear, ask for a documented connection between them.
Consider a fictional example: your packaging says LOT A124, while the report says A123. You can record that a report is available for another batch. You cannot transfer its results to your package simply because the product names match.
Sampling matters too. Who selected and submitted the material? Does the documentation explain how the sample represents the wider batch? Sampling and sample preparation can contribute to measurement uncertainty. A result for a submitted sample does not automatically establish the composition of the entire batch. (Eurachem, Measurement uncertainty arising from sampling, second edition, 2019.)
Step 2: distinguish the different dates
A report can contain manufacturing, sampling, receipt, analysis and issue dates. Each describes a different event. The date the laboratory issued the document is not necessarily the date it performed the tests.
Read the labels before deciding whether the timeline makes sense. If analysis appears to precede sample receipt, ask the issuer to explain the sequence. Do not invent a reason or reinterpret a field to make the dates fit.
Check the date format as well. A US report may put the month before the day. Write the month in words in your own notes when a numerical date could be read in two ways.
A recent report for the wrong batch does not resolve a missing connection. Equally, age alone does not determine whether an older report is useful for a particular purpose. Consider the tested material, the batch and the question being asked.
Keep the report reference, version and download date with your notes. If a revised document becomes available, preserve enough information to distinguish the versions. This helps you identify which findings and explanations informed your assessment.
Step 3: read values together with their units
A concentration needs a unit or a clearly stated reporting basis. Percentages, milligrams per gram, milligrams per millilitre and micrograms per kilogram describe quantities on different scales or bases. Check the unit before comparing the number.
For a mass percentage, 1% equals 10 mg/g. One gram in one hundred grams is equivalent to ten milligrams in each gram. A fictional result of 5% by mass therefore equals 50 mg/g. This is a calculation example, not a product result or usage recommendation.
Do not treat mg/g and mg/ml as interchangeable. Converting between them requires density information. Also check whether the value refers to dry material or the sample as received. Different reporting bases can make apparently similar figures unsuitable for direct comparison.
A total may be calculated from several components, while another row reports one individual substance. Read the formula or footnote before assuming they describe the same quantity.
Record four details together: value, unit, tested material and reporting basis. If one is missing, request clarification before calculating. A precise-looking result does not remove the uncertainty created by an unknown assumption.
Step 4: understand ND, LOD and LOQ
ND usually means ‘not detected’, but its precise meaning depends on the report’s legend. It does not prove absolute absence. Interpretation depends on the analytical method and the limit used by the laboratory.
The limit of detection, often abbreviated LOD, concerns distinguishing a substance’s signal from background. The limit of quantification, or LOQ, concerns the level at which an amount can be determined with suitable measurement performance. A detected signal below the LOQ does not automatically support a reliable exact concentration.
These limits relate to the analytical method and its application. Look for the applicable limit beside the relevant substance, rather than assuming one threshold applies throughout the document. (Eurachem, The Fitness for Purpose of Analytical Methods, third edition, 2025.)
Keep ‘not detected’ separate from ‘not tested’ and ‘not reported’. Codes such as NT and NR require the laboratory’s own explanation. A blank field is not a negative finding. If the legend is missing, ask what the entry means before using it in a comparison.
Step 5: check testing scope, limits and laboratory details
Read the individual test sections before accepting an overall summary. A composition panel does not automatically cover microbiological contamination, metals, pesticides or residual solvents. Look for the specific analyses relevant to your question.
If a result says ‘pass’, find the specification against which it was assessed. A pass is a conclusion against stated criteria, not proof that the measured substance is absent. The decision rule explains how the laboratory determines whether a result meets a requirement. (ILAC, Guidelines on Decision Rules and Statements of Conformity, 2019.)
Accreditation also has a defined scope. Check whether the relevant testing activity falls within the laboratory’s scope of accreditation. A logo does not establish that every analysis shown in a report is accredited. (ILAC, Guideline for describing Scopes of Accreditation, 2024.)
For a value close to an assessment boundary, ask how measurement uncertainty was considered. More decimal places do not automatically mean greater certainty. Avoid declaring one sample meaningfully different from another merely because the final digits differ.
An available COA as a historical reading example
Good Nature Vibe hosts this published analysis report from Infinite Chemical Analysis Labs for batch 230329CC. It records receipt and completion on 28 June 2023 and carries the wording ‘QA SAMPLE – INFORMATIONAL ONLY’.
The document is used here solely as a historical reading example. A connection to current stock has not been established. Its results should therefore not be attributed to a current package without supporting documentation.
The CBD row reports 99.16%. That figure belongs to the described sample, not every product carrying a similar name. The report states that cannabinoid concentrations are reported on a dry-weight basis.
The summary also lists testing dates in May 2023, earlier than the June receipt date. The sequence requires clarification from the issuer; the document does not explain it. Do not assume a cause for the discrepancy. (Infinite Chemical Analysis Labs, Certificate of Analysis for batch 230329CC, 2023.)
The practical lesson is to read identification, dates and limitations together. A downloadable file alone does not establish its relevance to a current package.
What does a COA not prove?
A COA does not demonstrate a personal health effect. Measuring composition is a different task from investigating whether a finished product produces a particular outcome in people.
It also does not establish that every possible substance was investigated. The testing scope is selected. If a topic is absent, record ‘not tested or not reported in this document’ rather than interpreting the gap as proof of absence or approval.
A report is not automatically proof of lawful sale, complete labelling or suitability for everyone. Those conclusions require their own supporting information. Assessment criteria used in one jurisdiction should not be presented as automatic approval in another.
Finally, a measurement at one point in time does not establish how a product changes throughout storage. For a conclusion about shelf life, look for evidence addressing that specific question. Keep what was measured separate from what the measurement is being used to suggest.
What should you do when information is missing?
If the relevant file is unavailable, ask for the document linked to your batch. Include the product name, the complete code and the particular point you need explained. A photograph of the code can help avoid transcription errors.
Before comparing two reports, write down the question you want answered. Are you checking an ingredient concentration, identifying the tested material or looking for a particular contamination test? A document can help with one question while leaving another unanswered.
Organise your notes under three headings: ‘confirmed’, ‘missing’ and ‘question’. Put only directly supported details in the first column. If an answer arrives later, keep it with the original file and note who supplied it.
Separate correspondence from the certificate itself. A supplier’s later explanation may help connect documents, but it should remain distinguishable from the laboratory’s original findings.
Use this checklist before writing your conclusion:
- The sample description matches the material being discussed.
- The batch matches, or its connection is documented.
- The dates are identified and any inconsistencies explained.
- Each relevant result has a unit and reporting basis.
- Untested items remain separate from non-detected substances.
- Assessment criteria and limitations have been considered.
Keep the original report beside your notes. Label your own summary clearly so it cannot be mistaken for a laboratory document. When asking a question, quote the relevant field without first guessing what an unfamiliar abbreviation means.
A useful message is: ‘My packaging shows batch A124, but the available report identifies A123. Which document relates to my batch?’ This gives the supplier a specific connection to investigate instead of asking generally whether testing exists.
If the answer remains incomplete, leave the gap visible. Receiving a document is not the same as receiving an explanation that supports your intended conclusion.
Frequently asked questions about a COA
Is a COA the same as a quality mark?
No. A COA records analyses and results. A quality mark operates under its own criteria and conditions. Read each according to what it actually covers.
Does ND mean a substance is completely absent?
No. ND generally means not detected within the limits of the method. Check the report’s legend and the applicable detection or reporting limit.
Can I use a COA for a different batch?
Not as direct evidence for your batch. Request the corresponding report or a documented explanation connecting the batch codes.
Is an older Certificate of Analysis always unusable?
No. Age alone does not determine that. Check the intended use and whether the product, batch, dates and circumstances support that use.
Does a COA prove that a product works?
No. Composition measurements and contamination tests do not establish a personal effect. A COA is not an efficacy study.
How Good Nature Vibe approaches documentation
Good Nature Vibe’s approach to quality emphasises relevant, available documentation and clear limits on conclusions. Reading a COA means checking what is present and identifying what still needs explanation.
Transparency also requires distinguishing a historical example from a report demonstrably linked to a current batch. This supports informed choices without suggesting more certainty than the document provides. (Good Nature Vibe, Our quality approach, no publication year stated; accessed 2026.)
Conclusion
A COA helps you ask precise questions about an identified sample, batch and set of results. Start with the connection to the product, then check dates, units, scope and limitations. Draw only conclusions supported by those details. Where information is missing, record the gap and request clarification rather than filling it with an assumption.
