How to Read a Certificate of Analysis

A certificate of analysis is the single most informative document in the supplement industry, and almost nobody reads one. It is also the check that underpins everything in our longevity supplements guide and our NMN benefits, side effects and dosage guide — because evidence for an ingredient is worthless if the capsule does not contain it.

Welzo publishes certificates for its own range — the NMN supplements collection, NMN Pro 1000 and the wider anti-ageing and longevity supplements category, all of which sit within the longevity stack framework. This guide teaches you to scrutinise ours as hard as anyone else's.

Most people do not know they can ask for one, and of those who do ask and receive one, most glance at a page of laboratory notation and take the existence of the document as reassurance in itself.

That is a mistake, because the document's existence proves very little. What matters is what it says, who produced it, and — critically — whether it describes the product in your hand or something else entirely. This guide takes you through a COA line by line, explains what each test means, tells you what good numbers look like, and lists the specific things that should make you put a product back down.

What a COA is — and what it is not

What it is

A certificate of analysis is a laboratory document reporting the results of tests performed on a specific, identified batch of material. It states what was tested, by which method, against what specification, and what result was obtained. It is signed or authorised by someone accountable for it.

In a properly documented supply chain, every production batch generates one.

What it is not

• It is not a safety approval. No regulator issues COAs. It is a private testing document, not a licence.
• It is not evidence that the product works. Confirming that a capsule contains 500 mg of an ingredient says nothing about whether that ingredient does anything. Efficacy is a separate question entirely — see longevity supplements that probably don't work.
• It is not a guarantee about the bottle you own unless the lot number matches.
• It is not automatically independent. A manufacturer can test its own products and issue its own certificate.

The distinction that catches almost everyone

If you take one thing from this article, take this.

There are two completely different documents that both get called a certificate of analysis:

• A raw material COA — issued by the ingredient supplier, testing the bulk powder before it becomes a product. It tells you the ingredient was good when it arrived at the factory.
• A finished product COA — testing the capsules or powder as sold. It tells you what is actually in the bottle.

Brands routinely publish the first and let customers assume it is the second. The raw material document is genuine, accurate and largely beside the point, because everything that can go wrong between bulk powder and finished capsule — incorrect blending, underdosing, degradation during processing, contamination from other production runs, loss of potency during shelf life — happens after that test.

Regulators make the same distinction. Under the EU contaminants framework — Commission Regulation (EU) 2023/915, which sets maximum levels for lead, cadmium, mercury, arsenic and other contaminants — the limits for food supplements apply to the ready-to-use supplement rather than to the individual raw materials [3,4,5]. Following EU exit, Great Britain operates its own assimilated contaminants legislation, while Northern Ireland applies EU rules under the Windsor Framework. The finished-product principle is common to both, which is the part that matters for reading a COA.

How to tell which one you are looking at

• A raw material COA usually names a chemical or ingredient — "NMN powder, 99% purity" — rather than a product name.
• It typically comes from a manufacturer in a producing country rather than a testing laboratory.
• It reports purity as a percentage of the material, not milligrams per capsule or per serving.
• It will not reference your product's brand name, bottle size or servings per container.

Both documents are useful. Only one answers the question you are actually asking.

Anatomy of a COA, section by section

1. The header — who is speaking

Look for the testing laboratory's name, address and accreditation number. If the header shows only the supplement brand, you are looking at a self-issued document. That is not automatically worthless, but it is a different category of evidence from an independent report.

2. Product identification and lot number

The product name should match what you bought, and there should be a batch or lot number. Check that number against the one printed on your bottle or pouch. If they do not match, the document describes a different batch and tells you nothing about yours.

This is the check most people skip and the one that matters most. A brand can publish a genuine COA for one excellent batch indefinitely while shipping something else.

3. Dates

Look for the date of manufacture, the date of analysis and, ideally, an expiry or best-before date. A COA dated three years ago attached to a currently shipping product is a warning sign. Some ingredients degrade — NMN converts toward nicotinamide with heat and moisture — so a recent test on a recent batch is meaningfully better than an old one.

4. The results table — specification versus result

This is the heart of the document, and it should have at least three columns:

• Test — what was measured.
• Specification — the acceptable range or limit.
• Result — what was actually found.

A fourth column stating the method used is a good sign. If a document gives results with no specification to compare them against, you cannot tell whether a number is good or bad — which is sometimes the point.

5. Identity and assay

Identity confirms the substance is what it claims to be. Assay (or potency) quantifies how much is present. Both matter: identity without assay tells you it is the right compound at an unknown dose; assay without identity tells you there is a lot of something.

For a finished product, look for the active stated per capsule or per serving, matching the label.

6. Heavy metals

Standard panel is lead, cadmium, mercury and arsenic. Each should have a specification and a result, and the result should be comfortably below the limit rather than scraping under it.

7. Microbiological testing

Typically includes total aerobic microbial count, yeast and mould, and specific pathogens — E. coli, Salmonella, Staphylococcus aureus. Pathogen results should read "absent", "not detected" or "negative" in the stated sample size.

8. Physical and organoleptic checks

Appearance, colour, odour, moisture content, particle size. These sound trivial but moisture content in particular predicts stability — a hygroscopic ingredient stored or packed wet will degrade faster.

9. Authorisation

A name, role and signature or electronic authorisation. An unsigned document nobody is accountable for is a weaker artefact than a signed one.

What a well-formed results table looks like

Set out below is the structure you should expect. The specific figures will differ by product; the shape should not.

What different product types need

• Synthesised single ingredients (NMN, creatine, TMG): identity, assay and purity are the priorities, with degradation products worth checking.
• Minerals (magnesium, zinc, iron): elemental content rather than compound weight, plus heavy metals.
• Botanicals and mushroom extracts (ashwagandha, shilajit, lion's mane): heavy metals, pesticide residues, microbiology and marker compound standardisation all matter more, because plant material concentrates whatever was in its soil.
• Oils (omega-3, algal oil): active fraction rather than total oil, plus oxidation markers such as peroxide value.
• Powders and probiotics: microbiology and moisture, plus viable count at end of shelf life for live cultures.

Understanding the test methods

You do not need to understand the chemistry. You do need to know that the method should be stated, and that some methods are more discriminating than others.

If a document reports a potency figure with no method stated, that is a gap. Method matters because a less specific technique can register a related compound as if it were the target — which is precisely how a degraded product can appear to pass.

What the numbers should look like

Potency

The result should meet or modestly exceed the label claim. Slight overages are normal and legitimate — manufacturers add a margin so the product still meets its claim at the end of shelf life after some degradation.

What should concern you is a result meaningfully below label claim, or a specification written so loosely that almost anything passes. A specification of "not less than 90% of label claim" is reasonable. A specification of "typical value" is not a specification at all.

Purity

For single-ingredient products, look for a stated purity percentage and note what the remainder is. For NMN, the usual impurity is nicotinamide, a degradation product — a high nicotinamide figure suggests age or poor storage.

Heavy metals

In the EU framework, maximum levels for contaminants in food including food supplements are set by Commission Regulation (EU) 2023/915, covering lead, cadmium, mercury, arsenic and other elements with category-specific limits [3]. Great Britain maintains its own assimilated version of this legislation, so the figures that apply depend on where the product is placed on the market. Rather than memorising limits, read the relationship between the two columns: a result far below its specification is reassuring, a result just under it less so.

Ingredients that concentrate metals from soil or water — botanicals, seaweed, algae, clays — warrant more attention here than synthesised compounds.

Microbiology

Counts should be within specification and pathogens absent. Powders and botanical materials carry more inherent microbiological risk than synthesised actives, so testing matters more for those.

Who tested it, and why that matters most

Independent versus in-house

A manufacturer testing its own product has a commercial interest in the result. That does not make in-house testing dishonest — most reputable manufacturers do it as routine quality control — but it is not the same as independent verification, and a brand describing self-testing as "third-party tested" is misdescribing it.

ISO/IEC 17025 accreditation

The relevant standard for testing laboratories is ISO/IEC 17025, which sets general requirements for the competence, impartiality and consistent operation of laboratories performing testing. In the UK, accreditation is granted by the United Kingdom Accreditation Service (UKAS) and covers staff qualifications and training, properly calibrated and maintained equipment, validated methods, and participation in proficiency testing schemes that compare a laboratory's results against others [6,7].

Crucially, accreditation is granted for specific tests, not for a laboratory as a whole — a lab can be accredited for heavy metals and not for the assay method used on your product. A COA from an accredited laboratory that states its accreditation number lets you check the scope.

The hierarchy, in order of strength

1. Finished product tested by an ISO 17025-accredited independent laboratory, batch-specific, recent, method stated.
2. Finished product tested by an independent laboratory without stated accreditation.
3. Finished product tested in-house by the manufacturer.
4. Raw material COA from the ingredient supplier.
5. A claim of "lab tested" with no document at all.

Most products in this market sit at level four or five.

Twelve red flags

Reading a COA in sixty seconds

You will not analyse every document in full. Here is the fast version, in order:

1. Whose name is at the top? A laboratory, or the brand?
2. Is there a lot number, and does it match my bottle?
3. What is the date of analysis?
4. Does it describe the finished product or a raw ingredient?
5. Does the assay result meet the label claim?
6. Are heavy metals and microbiology both present and within specification?

Six questions. If a document answers all six well, you are buying from someone doing this properly. If it fails two or more, the price no longer matters — see why supplements are expensive for why testing is the line item worth paying for, and third-party tested supplements for the wider context.

A note on bundles and multi-ingredient products

If you buy a bundle or a multi-ingredient formula, one certificate does not cover everything. A bundle of three separate products should have three sets of documentation, each with its own lot number. A single formula containing several actives should show assay results for each of them — and if it only reports one, ask why the others were not measured.

This is also where proprietary blends become a practical problem rather than a theoretical one: if the label does not disclose per-ingredient amounts, a certificate cannot verify them against anything. See supplement fillers explained.

What to do if a brand won't provide one

Ask directly, by email, for the batch-specific certificate of analysis for the lot number on your product, from an independent laboratory. A reputable brand will send it or point you to where it is published.

Responses that should concern you: refusal on grounds of commercial confidentiality (the COA is not a formulation), sending a raw material document instead, sending an undated or lot-free document, or simply not replying. Any of these tells you what you needed to know.

If you have already bought and cannot get documentation, you have UK consumer rights around goods being as described. If you believe a product is mislabelled or unsafe, you can report it to your local authority Trading Standards service via the Citizens Advice consumer helpline.

Applying this to a purchase decision

Documentation quality should sit alongside dose and price in your comparison, not after them. A cheaper product with a published batch COA is a better purchase than a premium one without — and given that independent testing has found products containing no detectable active at all [1,2], it is not a close call.

Practical buying guides across the range: the minimal supplement stack, the cheap longevity stack, how much a longevity stack costs, is NMN a scam?, the cheapest NMN in the UK and where to buy NMN in the UK. Our own range is in Welzo Ultra Purity, with our ingredient quality and purity certifications set out separately.

Where documentation matters most: single high-cost actives such as NMN Pro 1000 and Trans-Resveratrol, where you cannot judge content by taste or appearance; botanical products such as Shilajit complex, where contamination risk is inherently higher; and anything bought from a marketplace listing rather than an accountable retailer. Related reading on evaluating specific compounds: does resveratrol actually work?, the NAD supplement guide, shilajit UK guide, vegan longevity supplements and how long supplements take to work.

When to speak to a doctor

Speak to your GP or pharmacist before starting any supplement if you take prescription medication, have an existing medical condition, or are pregnant or breastfeeding. A clean certificate of analysis confirms a product contains what it claims — it says nothing about whether that product is appropriate for you.

If you have taken a supplement and experienced an adverse reaction, stop taking it and seek medical advice. Adverse reactions to food supplements can be reported to the Food Standards Agency, and keeping the packaging and batch number will materially help any investigation.

Frequently asked questions

What is a certificate of analysis for a supplement?

It is a laboratory document reporting test results for a specific batch of product, stating what was tested, by what method, against what specification, and what was found. It typically covers identity, potency, heavy metals and microbiological contamination. It is a private testing document, not a regulatory approval.

How do I know if a COA is genuine?

Check that a testing laboratory is named with contact details and, ideally, an accreditation number you can verify. Check that the batch number matches your product. Check the date is recent. Check that specifications appear alongside results. A low-resolution image with no laboratory named and no lot number is not a usable document.

What is the difference between a raw material COA and a finished product COA?

A raw material COA tests the bulk ingredient before manufacturing; a finished product COA tests the capsules or powder as sold. Brands frequently publish the former while customers assume the latter. Only the finished product document tells you what is in your bottle — and contaminant limits for supplements, under both the EU regulation and the assimilated Great Britain legislation, apply to the ready-to-use product rather than the individual raw materials [3].

What does ISO 17025 mean on a lab report?

ISO/IEC 17025 sets general requirements for the competence, impartiality and consistent operation of testing laboratories, with UK accreditation granted by UKAS. It covers staff competence, equipment calibration, method validation and proficiency testing against other laboratories [6,7]. Accreditation is granted for specific tests rather than the laboratory as a whole, so an accreditation number lets you check the scope.

What heavy metal levels are acceptable in a supplement?

Maximum levels for lead, cadmium, mercury, arsenic and other contaminants in food including supplements are set in the EU by Commission Regulation (EU) 2023/915, with category-specific limits; Great Britain operates its own assimilated version [3]. Rather than memorising figures, read the relationship between the specification and result columns — a result comfortably below its limit is reassuring, one just under it less so. Botanicals, seaweed and algae warrant closer attention than synthesised compounds.

Why does my COA show more active ingredient than the label says?

Modest overages are normal and legitimate. Manufacturers add a margin so that the product still meets its label claim at the end of shelf life, after some degradation. A result meaningfully below label claim is the concerning direction, not above.

Should the batch number on the COA match my bottle?

Yes, and this is the check most people skip. A certificate describes one specific batch. If the lot number does not match, the document tells you about a different production run — potentially an excellent one, published indefinitely while something else ships.

Does a COA mean the supplement works?

No. It confirms content and purity, not efficacy. A product can be verifiably exactly what it claims and still have no meaningful evidence behind the ingredient — NMN is a good example, where the compound reliably raises NAD⁺ but a meta-analysis of eight randomised trials found no significant benefit on glucose or lipid markers [8]. Those are separate questions — see longevity supplements that probably don't work and does resveratrol actually work?

What if a brand says its COA is confidential?

Treat that as an answer. A certificate of analysis reports test results for a batch — it does not disclose a formulation or trade secret. Confidentiality is not a credible reason to withhold one from a customer who has bought the product.

Are supplements tested before they are sold in the UK?

Not for efficacy. Food supplements are regulated as foods rather than medicines, so they must be safe and accurately labelled but do not undergo pre-market approval demonstrating they work. Contaminant limits apply under the relevant legislation for the market in question, and businesses must comply [3], but batch testing by an independent laboratory is a voluntary quality measure rather than a legal requirement — which is exactly why its presence is so informative.

References

1. NUS Medicine. Significant discrepancies between actual and labelled amount of anti-ageing ingredients in supplements. National University of Singapore, Yong Loo Lin School of Medicine, 2024. https://medicine.nus.edu.sg/news/significant-discrepancies-between-actual-and-labelled-amount-of-anti-ageing-ingredients-in-supplements-nus-medicine-study/
2. Sandalova E, et al. Testing the amount of nicotinamide mononucleotide and urolithin A as compared to the label claim. GeroScience. 2024;46:5075–5083. Reported in NutraIngredients: https://www.nutraingredients.com/Article/2024/08/23/Many-NMN-and-Urolithin-A-supplements-failing-to-match-label-claims/
3. European Commission. Contaminants — Legislation: Commission Regulation (EU) 2023/915 on maximum levels for certain contaminants in food. https://food.ec.europa.eu/food-safety/chemical-safety/contaminants/legislation_en
4. GBA Group. Contaminants in food supplements and botanical raw materials. 2024. https://www.gba-group.com/en/blog/food/contaminants-in-food-supplements-and-botanical-raw-materials/
5. Measurlabs. Food contaminant testing by EU regulations. https://measurlabs.com/blog/eu-regulations-on-food-contaminants/
6. United Kingdom Accreditation Service. Laboratory accreditation — ISO/IEC 17025. https://www.ukas.com/accreditation/standards/laboratory-accreditation/
7. United Kingdom Accreditation Service. Spotlight on accreditation standards: the importance of ISO/IEC 17025 for laboratory testing and calibration. https://www.ukas.com/resources/resources/soas-17025/
8. Effects of Nicotinamide Mononucleotide on Glucose and Lipid Metabolism in Adults: A Systematic Review and Meta-analysis of Randomised Controlled Trials. Current Diabetes Reports. 2024. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11557618/

All product photography in this article is owned by Welzo and served from the Welzo media library. Regulatory positions correct at the time of publication and subject to change.