Third-Party Testing Explained: What USP, NSF, and ConsumerLab Certifications Actually Verify

Key Takeaways

Certification programs that audit the factory and retest the product over time (USP, NSF) give more assurance than a single lab report. Sport programs add screening for drugs banned in competition, which matters because one international study found undeclared anabolic steroids in 14.8% of 634 supplements that did not list any hormones ( Geyer et al., 2004, Int J Sports Med). ConsumerLab buys products off the shelf, so manufacturers cannot hand-pick samples. For any seal, look up the product on the certifier's own site before trusting the logo, and when there is no seal, ask for a batch Certificate of Analysis with numbers. The supplement label guide covers the other lines to check.

Third-party testing explained: what problem is it solving?

It fills a gap in U.S. law. Supplements do not need FDA approval before they are sold, and manufacturers vouch for their own labels. Independent testing is a voluntary way for a company to prove its label is accurate and for a buyer to check that claim.

The Dietary Supplement Health and Education Act of 1994 put supplements under food law. Manufacturers must follow current Good Manufacturing Practice rules in 21 CFR Part 111, but nobody outside the company tests a product before it reaches the shelf. FDA enforcement mostly happens afterward, through inspections, warning letters and recalls.

Researchers have argued for years that this after-the-fact system moves slowly. A 2015 analysis in the American Journal of Public Health described FDA action against adulterated supplements as too limited and too late to protect consumers ( Starr, 2015, Am J Public Health). The data on hidden drugs supports that concern. From 2007 to 2016 the FDA identified 776 adulterated supplements from 146 companies, and 157 of them contained more than one unapproved ingredient ( Tucker et al., 2018, JAMA Netw Open). Problems are not limited to hidden drugs. A review of 51 Ayurvedic herbal products sold in the United States found labeling violations on 61% of them ( Jordan et al., 2023, Food Control), and an older analysis of 12 over-the-counter steroid-precursor products found that 11 contained less than the labeled amount ( Green et al., 2001, Clin J Sport Med). Third-party testing does not fix the law, but it gives careful companies a way to separate themselves from that part of the market.

What exactly does a lab check?

Most programs cover three questions: whether each ingredient is correctly identified, whether the measured amount falls within an allowed range of the label claim, and whether contaminants are below set limits. Some programs add dissolution testing, banned-substance screening and a factory audit.

Identity testing uses methods such as high-performance liquid chromatography, mass spectrometry and, for plants, microscopy or DNA analysis. It catches a cheaper ingredient swapped in for an expensive one. Botanicals are where swaps are most common. A DNA barcoding study found that 59% of 44 herbal products contained plant species not listed on the label, and substitution occurred in 30 of the 44 ( Newmaster et al., 2013, BMC Med).

Quantity testing compares the measured amount with the label. Each program sets its own acceptable range, and ranges differ by ingredient type, so a botanical marker compound may have a wider allowance than a vitamin. Contaminant testing typically covers lead, arsenic, cadmium and mercury, microbial counts such as yeast, mold, E. coli and Salmonella, and in some programs pesticide and solvent residues.

Identity and amount can be hard to measure even for well-equipped labs. Validated methods for complex botanical mixtures are still being developed, which is one reason the NIH and federal partners have invested in reference materials and analytically checked label databases ( Betz et al., 2018, J Nutr).

How do USP, NSF, Informed Sport and ConsumerLab differ?

USP Verified and NSF certification are manufacturer-funded programs that combine product testing with facility audits and ongoing checks. NSF Certified for Sport and Informed Sport add screening for substances banned in sport. ConsumerLab buys products at retail and publishes results, but it does not audit factories.

USP is a nonprofit that also writes the official quality standards used for medicines in the United States. Its supplement program checks that the product contains the listed ingredients in the declared amounts, does not exceed limits for specified contaminants, breaks down properly, and was made under good manufacturing practice. NSF, also nonprofit, runs a contents-certification program built on its dietary supplement standard and a sport program layered on top of it.

Informed Sport, run by the lab company LGC, focuses on prohibited substances and tests batches before release. ConsumerLab is a for-profit company with a different model: it chooses products, buys them from stores and websites, tests them, and publishes pass/fail results with measured amounts. Companies can pay for a "voluntary certification" track, but the core reviews do not depend on manufacturer participation.

Is a sport certification worth paying attention to if I am not an athlete?

Often, yes. Banned-substance screens look for stimulants, anabolic agents and other drugs that standard contaminant panels may not include. Those compounds turn up as undeclared ingredients in categories that ordinary consumers buy too, such as pre-workouts, weight-loss products and "testosterone support" formulas.

The international study by Geyer and colleagues bought 634 non-hormonal supplements from 215 suppliers in 13 countries. After excluding samples that could not be analyzed reliably, 94 products (14.8%) contained anabolic steroids that were not on the label ( Geyer et al., 2004, Int J Sports Med). Most buyers were not trying to take steroids. They were buying protein, vitamins and amino acids.

If you take creatine, protein or any performance product, a sport certification is a reasonable tiebreaker even if you will never face a drug test. For vitamins and simple minerals, a general certification is usually enough.

What does a Certificate of Analysis tell me?

A Certificate of Analysis (COA) reports test results for one specific batch. A useful one names the testing lab, gives the batch or lot number, lists each test with its method, specification and actual result, and is dated. Numbers are more informative than the word "pass."

Start by matching the lot number on the COA with the one printed on your bottle. A COA for a different batch tells you about that batch, not yours. Next, check whether the lab is accredited to ISO/IEC 17025, the standard for testing labs. Then read the results. "Lead: 0.12 ppm (limit 0.5 ppm)" is clear. "Heavy metals: complies" is not, because you cannot see how close the result came to the limit or which metals were tested.

Also check whether the COA covers the finished product or only a raw ingredient. Many brands share the supplier's certificate for the powder they bought, which is useful but does not account for anything that happened during blending and encapsulation. For an example of why the scope of a heavy-metal panel matters, see our piece on thallium testing in shilajit.

Does GMP certification mean a product was tested?

Not on its own. Good manufacturing practice rules cover how a facility runs: identity checks on incoming materials, process controls, cleaning, record-keeping and complaint handling. A "GMP" or "made in a GMP facility" statement describes the factory, not the contents of your particular bottle.

GMP compliance is required by law for every U.S. supplement manufacturer, so the phrase on a label is closer to a baseline than a mark of distinction. Some companies hire outside auditors to certify their GMP systems, which adds credibility, but the audit still reviews procedures rather than testing finished products. The strongest position is a GMP-audited facility plus independent testing of the finished product, which is what USP and NSF programs combine.

What happens when a product fails?

In manufacturer-funded programs, a failing product loses the right to use the seal until the company fixes the problem and passes again. ConsumerLab publishes failures openly. A failed private test does not by itself trigger an FDA recall, although a certifier may report serious safety findings.

The gap between a failed test and a product leaving the market can be long. When researchers bought supplements at least six months after FDA recalls for hidden drugs, about two-thirds still contained a banned pharmaceutical ( Cohen et al., 2014, JAMA). Some products named in repeated FDA warnings came back with new hidden ingredients the second or third time ( Tucker et al., 2018, JAMA Netw Open). A current certification listing is evidence that a product passed recently. It does not follow that the rest of the market is policed with the same speed.

Why do certified products often cost more?

Certification adds recurring costs: program fees, audits, routine batch testing, documentation and, in many cases, more expensive qualified raw materials. Those costs are built into the price. The fees vary by program and by the number of products, and the certifiers do not publish a single price list.

A fair way to judge the premium is cost per verified dose. If a certified vitamin D costs a few cents more per day than an unverified one, you are paying that difference for independent confirmation of the dose and contaminant limits. For a supplement taken daily for years, that is a small amount for a known quantity. The comparison changes when the uncertified product has published, recent batch COAs from an accredited lab; then the gap in assurance is smaller.

Small brands often cannot afford full certification programs. That does not mean their products are poor, but it does shift the burden of proof to them. A small brand that posts lot-specific COAs with numeric results is being more transparent than a large brand with a slogan and no data.

How can I verify a seal in under a minute?

Go to the certifier's own website, search the brand and product name, and confirm that the exact product and format (capsule, powder, gummy) is listed. If it is not, treat the logo on the bottle as unverified and contact the certifier or the brand.

Use the links in the comparison table above. Two details trip people up. First, certification usually applies to specific products, not to a whole brand, so one certified fish oil does not cover the same company's multivitamin. Second, reformulated products may lose or need to renew certification, so an older bottle design with a seal may not match what is being sold now.

When there is no seal, a short email asking "Can you send the Certificate of Analysis for lot number X, including heavy metals and microbial results, and the name of the testing lab?" tells you a lot. A prompt reply with a full report is a good sign. Silence or a one-line "all our products are tested" answer is not.

What can third-party testing not tell me?

It cannot tell you whether a product works, whether the dose is right for you, how well the ingredient is absorbed in your body, or whether it interacts with your medicines. Those questions depend on clinical research and your own health situation.

A certified magnesium oxide tablet can contain exactly 400 mg as labeled and still deliver less usable magnesium than a smaller dose of a better-absorbed form. A certified fish oil contains the labeled EPA and DHA, but the certification makes no promise about heart outcomes. A certified herbal extract matches its label, yet the amount may be far below what was used in human studies.

Testing is also a snapshot of batches sampled at a point in time. Ongoing programs reduce that limitation with periodic retesting, but no system samples every bottle. Use third-party testing as one filter among several: dose and form compared with research, label transparency, the company's regulatory history, and advice from a pharmacist or clinician for anything you take with prescription drugs. Emergency department data show why the last step matters, with an estimated 23,005 supplement-related visits a year in the United States ( Geller et al., 2015, N Engl J Med).

Which supplement categories most need independent testing?

Testing matters most where problems have been documented most often: sexual-performance, weight-loss and muscle-building products, which dominate FDA findings of hidden drugs; herbal products, where species substitution is common; and anything sourced from soil or sea, where heavy metals and other contaminants can accumulate.

The FDA data are lopsided. Of 776 adulterated supplements identified between 2007 and 2016, 45.5% were sold for sexual enhancement, 40.9% for weight loss and 11.9% for muscle building ( Tucker et al., 2018, JAMA Netw Open). The hidden ingredients were prescription drugs or drugs that had been withdrawn, which is a different kind of risk from a vitamin that is 10% under its label claim. In these categories, a sport certification or similarly rigorous screen is the minimum we would look for.

Botanicals form the second group. Plants are easy to confuse, harvests vary, and powders look alike. The DNA barcoding work by Newmaster and colleagues found substitution in most of the herbal products tested and contaminants in a third of those that were authentic ( Newmaster et al., 2013, BMC Med). Ask for identity testing and, where possible, a stated marker compound.

The third group is ingredients that concentrate what is in their environment: fish oils, seaweed products, mineral-rich resins such as shilajit, and mushrooms grown on wood or soil. For these, the heavy-metal section of the COA is the part to read first. Products sold in emergency-department case reports skew toward the first group; adverse event visits most often involved weight-loss and energy products among adults ( Geller et al., 2015, N Engl J Med).

Reading a Certificate of Analysis: a worked example

A good COA is a table with four columns: the test, the method, the acceptance limit and the result. The example below is a made-up report for a magnesium glycinate capsule, laid out the way most accredited labs present results, with notes on how to read each line.

The limits in this example are illustrative; real limits depend on the product, the serving size and the standard the lab is applying. What matters is the structure. Every line names a method, so you know how it was measured. Every result has a number or a defined non-detect, so you can see the margin. The batch number at the top of the report should match your bottle, and the lab's name and accreditation should appear in the header or footer.

Compare that with a one-line statement such as "Heavy metals: Pass." You cannot tell which metals were tested, which limit was used, or whether the result sat just under it. If a brand sends you a report like that, it is reasonable to ask for the full version.

Lab terms that appear on test reports

A handful of abbreviations cover most of what you will see on a supplement COA. Knowing them takes a few minutes and makes it much easier to tell a thorough report from a thin one.

A result of "ND" or "< LOQ" is not the same as zero. It means the amount, if any, is below what that method can measure. That is normally a good outcome. It becomes less reassuring if the method's detection limit is high relative to the safety limit, which is why some careful reports list the LOQ beside each result.

Testing claims that should make you skeptical

Watch for "FDA approved" or "FDA certified" on a supplement, seals from organizations you cannot find, "made in an FDA-registered facility" presented as a quality badge, and offers of a COA that never arrives. Each of these borrows the look of verification without the substance.

The FDA does not approve dietary supplements, so "FDA approved" on a supplement label is false. Facility registration with the FDA is a legal requirement for food and supplement facilities and does not mean the FDA has inspected or endorsed anything. Brands sometimes print it next to a logo that looks official, which can give the wrong impression.

Seals are easy to design. If you cannot find the certifying organization online, or you can find it but the product is not in its public listing, treat the seal as marketing. Generic "quality tested" badges with no organization name fall in the same group. Lastly, a COA should exist before you ask for it. A company that promises one "on request" and then cannot produce a lot-specific report has told you something about its testing.

How did U.S. supplement testing end up voluntary?

DSHEA, passed in 1994, set up the current system: supplements need no pre-market approval, and the FDA must show a product is unsafe or mislabeled after it is sold. Federal GMP rules specific to supplements followed in 2007. Third-party programs grew to fill the space between those rules and what buyers wanted to know.

When DSHEA passed, the market was a fraction of its current size, and the internet barely existed as a retail channel. The law put the burden of label accuracy on companies and gave the FDA tools that mostly work after a problem appears. The supplement GMP rule in 21 CFR Part 111 then required manufacturers to verify the identity of ingredients and establish specifications for their products, which raised the floor but left testing done in-house.

Since then, sales have moved heavily online, where products from many countries compete side by side and new brands can launch in weeks. Critics argue the post-market model has not kept up ( Starr, 2015, Am J Public Health). Third-party certification, retailer requirements and published batch reports are the market's answer, and none of them is mandatory.

What each type of contaminant panel covers

"Contaminant tested" can mean very different things. The core panels are heavy metals, microbes, pesticides and residual solvents. Oils may also be tested for oxidation, and some products are screened for gluten, allergens or banned drugs. Ask which panels were run, because a report can be accurate and still leave out the test you care about.

Heavy metals. Lead, arsenic, cadmium and mercury are the standard four. They matter most for botanicals, minerals, seaweed, mushrooms, protein powders and resins. Some specialized panels add thallium or other elements for specific ingredients.

Microbes. Total aerobic count, yeast and mold, and specific pathogens such as E. coli, Salmonella and Staphylococcus aureus. These matter most for raw plant powders and anything that is not heat-treated.

Pesticides and solvents. Pesticide screens look for residues from farming; solvent screens look for traces of the chemicals used to make extracts. Both are more relevant to botanical extracts than to synthetic vitamins.

Oxidation. Fish, krill and algal oils can turn rancid. Peroxide value, anisidine value and their combined TOTOX score show how oxidized an oil is. A fish oil COA without these numbers is missing the test most specific to the product.

Banned substances. Sport programs screen for stimulants, anabolic agents, diuretics and other compounds on anti-doping lists. This panel is where hidden-drug problems are caught, and it is rarely included in a standard COA unless the product is sport certified.

What a credible brand testing program looks like

A brand with a serious testing program usually tests raw materials on arrival, tests finished batches before release, uses an accredited outside lab for at least the finished-product tests, and makes lot-specific results available. Certification by USP, NSF or a sport program sits on top of that.

Raw-material testing confirms that what the supplier shipped is what was ordered. Finished-product testing confirms that blending, encapsulation and packaging did not change anything. Using an independent lab removes the conflict of a company grading its own work. Posting results, or sending them on request within a day or two, shows the company expects scrutiny.

You can check most of this from the outside. Look for a quality or testing page on the brand's website that names the lab, explains which tests are run on every batch, and lets you look up results by lot number. Then compare what it says with what a sample COA actually shows. If the page promises heavy-metal testing and the report lists only microbes, the page is not accurate.

Retailers have also started to require documentation from sellers. Amazon introduced quality documentation requirements for supplements several years ago ( Crawford et al., 2022, JAMA Netw Open), but the same study found most of the products it bought there did not match their labels, which shows the limits of paperwork without independent testing.

Frequently asked questions

Is third-party testing required by law?

No. Manufacturers must follow GMP rules and keep their labels truthful, but independent testing and certification are voluntary.

Can a supplement be good quality without a seal?

Yes. Some companies test thoroughly but skip certification because of cost. Without a seal, ask for recent, lot-specific COAs from an accredited lab.

Does "lab tested" on a label mean third-party tested?

Not necessarily. It can mean the company tested the product in its own lab or tested only the raw ingredients. Look for the name of an independent lab or certifier.

Which seal is best for general use?

USP Verified and NSF certification both include facility audits and ongoing testing, which makes them strong choices for everyday vitamins and minerals.

Which seal should athletes look for?

NSF Certified for Sport or Informed Sport, since both screen for substances on anti-doping lists. Check your sport's governing body for its preferred program.

How often are certified products retested?

It depends on the program. Ongoing programs retest periodically or by batch, which is why you should check the certifier's current listing rather than a seal from years ago.

Are mushroom and herbal products harder to test?

Yes. Botanicals contain many compounds and vary naturally, so identity and potency testing needs specialized methods. For chaga and other mushrooms, ask whether the COA covers heavy metals and the specific marker compound, such as beta-glucans.

References

1. 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;5(8):e2226040. PubMed 35947382
2. Geyer H, Parr MK, Mareck U, et al. Analysis of non-hormonal nutritional supplements for anabolic-androgenic steroids - results of an international study. Int J Sports Med. 2004;25(2):124-9. PubMed 14986195
3. Starr RR. Too little, too late: ineffective regulation of dietary supplements in the United States. Am J Public Health. 2015;105(3):478-85. PubMed 25602879
4. Tucker J, Fischer T, Upjohn L, et al. Unapproved Pharmaceutical Ingredients Included in Dietary Supplements Associated With US Food and Drug Administration Warnings. JAMA Netw Open. 2018;1(6):e183337. PubMed 30646238
5. Jordan CR, Harris CM, Miranda MI, et al. Labeling compliance and online claims for Ayurvedic herbal supplements on the U.S. market associated with the purported treatment of COVID-19. Food Control. 2023;148:109673. PubMed 36778101
6. Green GA, Catlin DH, Starcevic B. Analysis of over-the-counter dietary supplements. Clin J Sport Med. 2001;11(4):254-9. PubMed 11753063
7. Newmaster SG, Grguric M, Shanmughanandhan D, et al. DNA barcoding detects contamination and substitution in North American herbal products. BMC Med. 2013;11:222. PubMed 24120035
8. Betz JM, Rimmer CA, Saldanha LG, et al. Challenges in Developing Analytically Validated Laboratory-Derived Dietary Supplement Databases. J Nutr. 2018;148(suppl_2):1406S-1412S. PubMed 31505676
9. Cohen PA, Maller G, DeSouza R, et al. Presence of banned drugs in dietary supplements following FDA recalls. JAMA. 2014;312(16):1691-3. PubMed 25335153
10. Geller AI, Shehab N, Weidle NJ, et al. Emergency Department Visits for Adverse Events Related to Dietary Supplements. N Engl J Med. 2015;373(16):1531-40. PubMed 26465986
11. 21 CFR Part 111 (official source). www.ecfr.gov
12. nsfsport.com (official source). www.nsfsport.com
13. Informed Sport search (official source). sport.wetestyoutrust.com
14. consumerlab.com (official source). www.consumerlab.com