The short version
- A certificate of analysis covers one lot of one product on one date, and reports identity, potency and contaminant results separately — a clean heavy-metals screen says nothing about whether the active is at label dose.
- Of 30 immune-support supplements analysed by mass spectrometry in a 2022 case series, 17 had inaccurate labels, 13 listed undetectable ingredients, nine contained undeclared substances, and none carried a third-party certification seal.
- An analysis of FDA warnings found 776 adulterated dietary supplements between 2007 and 2016 across 146 companies, concentrated in sexual enhancement (45.5%), weight loss (40.9%) and muscle building (11.9%) products.
- A certificate of analysis reports composition, not efficacy: a verified batch of an ingredient with no human trials behind it is still an ingredient with no human trials behind it.
- Certificates are issued at manufacture, so they say nothing about oxidation or moisture damage that happens on a shelf afterwards.
Ask a supplement brand whether its products are tested and you will almost never get a no. Ask who tested them, which batch, and for what, and the answers thin out fast. That gap is the whole subject of this article.
A certificate of analysis is a lab report on one specific lot of one specific product. It typically states an identity result, a potency result for the actives, and a set of contaminant screens — heavy metals, microbial counts, sometimes solvents or pesticides. Read properly it is the single most useful document a brand can hand you. Read as a badge, it proves less than people assume.
Why the document matters at all
Because labels are wrong often enough to be a live risk rather than a theoretical one. In a 2022 case series in JAMA Network Open, researchers bought 30 immune-support supplements from a major online marketplace and analysed them by mass spectrometry. Thirteen had accurate labels. Of the 17 that did not, 13 listed ingredients that could not be detected at all and nine contained substances that were never declared. Not one of the 30 carried a third-party certification seal.
The harder end of the same problem is adulteration. An analysis of the FDA's tainted-products database found 776 adulterated dietary supplements identified between 2007 and 2016, implicating 146 companies. They clustered in three categories: sexual enhancement (45.5%), weight loss (40.9%) and muscle building (11.9%). The most common hidden ingredients were sildenafil, sibutramine, and synthetic steroids respectively. Of 28 products named in more than one warning, 19 turned up with a new unapproved ingredient the second or third time.
So the case for testing is not paranoia. It is the base rate.
The four questions that separate a real COA from a marketing asset
1. Who ran it? "Third party" means a lab with no ownership stake in the outcome. An in-house QC report can be perfectly competent and still is not third-party. Look for a named, accredited laboratory rather than the phrase "independently tested" with no lab on the page.
2. Which lot, and does it match the bottle in your hand? This is the one almost nobody checks. A COA covers a lot number and a manufacture date. If the certificate on the website is three years old and your bottle's lot code does not appear on it, you are reading someone else's test. Brands that are serious about this let you look up your specific lot.
3. What was it tested for? Identity, potency and contaminants are three different tests. A heavy-metals screen says nothing about whether the active is at label dose. A potency assay says nothing about lead. Many certificates cover one and imply the other.
4. What are the limits, and is a detection actually a problem? This cuts against the alarmist reading. A risk assessment of heavy metals reported in protein powders modelled arsenic, cadmium, mercury and lead at one and three servings a day and calculated hazard indices below 1 across the products examined, with background exposure the main contributor to modelled blood lead. That analysis came from a consulting firm and is a model rather than a measurement in people, so treat the number as directional. The point stands: detected is not the same as unsafe, and a certificate that reports a trace figure is doing its job, not confessing.
What a COA still cannot tell you
It cannot tell you the ingredient works. This is the substitution people make most often — a clean certificate feels like evidence of efficacy, and it is evidence of composition. A perfectly tested batch of something with no human trials behind it is a perfectly tested batch of something with no human trials behind it.
It cannot tell you the dose is the studied dose. That is a label-reading job, not a lab job.
It cannot tell you what happened to the bottle after the lab saw it. Certificates are issued at manufacture. Oxidation and moisture do their work later, which is why expiry dates and how supplements are stored are a separate question from purity at packing.
And it cannot tell you the next lot will look the same. That is what recurring testing is for, and it is the expensive part.
What we'd actually tell you
Treat certification as a floor, not a finish. If a brand will not name the lab, will not let you match a lot number, or publishes one certificate for an entire product line, ask why — including if the brand is us.
Then apply the second filter, which is harder and matters more: a verified batch of something you did not need is still money spent. Purity is a reason to trust a purchase. It is not a reason to make one, and when you cut your stack the items you kept because they are clean rather than because they do anything should go first.
Good questions
What is a certificate of analysis on a supplement?
It is a laboratory report on one specific production lot. It normally covers identity (is this the ingredient claimed), potency (how much is actually in there) and contaminants (heavy metals, microbes, sometimes solvents or pesticides). It is issued at manufacture and applies to that lot only, not to the product line as a whole.
Does third-party tested mean the supplement works?
No. Testing verifies composition, not effect. A brand can produce a flawless certificate for an ingredient that has never shown a benefit in a human trial. Use the certificate to decide whether you can trust what is in the bottle, and use the research to decide whether what is in the bottle is worth taking.
How do I check a certificate of analysis myself?
Find the lot number printed on your bottle and confirm it appears on the certificate. Check the laboratory is named and accredited, and is not owned by the brand. Check which tests were actually run, because identity, potency and contaminant screening are separate. If the certificate is undated, unlotted, or covers an entire product range, it is marketing.
Should I worry about heavy metals in my supplements?
Be aware rather than alarmed. Trace metals are detectable in many plant-derived and mineral products because they are detectable in soil. One modelled risk assessment of protein powders calculated hazard indices below 1 at up to three servings a day, with background exposure the main driver of estimated blood lead. That is a model, not a measurement in people. The meaningful question is whether a brand tests and publishes at all.
Is an in-house lab report worthless?
Not worthless, just not independent. A competent manufacturer running its own release testing is doing real work, and most reputable products depend on it. The difference is that a third-party lab has no stake in the result. If a brand only has in-house data, it should say so plainly rather than using the phrase independently tested.
Do certification seals like a sport certification cover everything?
No, and this trips up athletes in particular. Banned-substance certification programmes screen for a list of prohibited compounds in a tested lot. They are not a general statement of purity, they do not confirm the label dose, and they apply to certified lots rather than to every batch a brand ever makes. Check what the specific seal covers before you rely on it.
Sources
- Crawford C, Avula B, Lindsey AT, et al. Analysis of Select Dietary Supplement Products Marketed to Support or Boost the Immune System. JAMA Netw Open, 2022. View study
- Tucker J, Fischer T, Upjohn L, Mazzera D, Kumar M. Unapproved Pharmaceutical Ingredients Included in Dietary Supplements Associated With US Food and Drug Administration Warnings. JAMA Netw Open, 2018. View study
- Bandara SB, Towle KM, Monnot AD. A human health risk assessment of heavy metal ingestion among consumers of protein powder supplements. Toxicol Rep, 2020. View study
- Geller AI, Shehab N, Weidle NJ, et al. Emergency Department Visits for Adverse Events Related to Dietary Supplements. N Engl J Med, 2015. View study
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
