A 2024 study run jointly by the Uniformed Services University and the University of Mississippi tested 44 weight-loss products bought at or near military installations. Twenty-seven of the 44 listed ingredients on the label that the laboratory could not detect at all, with between 6 percent and 50 percent of a given label’s ingredients simply missing.
Not one of the 44 carried a seal from any of the established third-party certification programmes. Two carried seals that looked like certifications and were not. One said the product was quality tested. The other promised ironclad quality and guarantee. No third-party organisation stood behind either mark.
That is the shape of the problem. A seal is a claim, exactly like a claim about what a product does, and US law has treated it that way for years. What the law does not do is check seals before they go on a bottle. Nothing screens a badge, nothing registers one, and a graphic that reads like a certification can be designed in an afternoon.
This article is about the marks and phrases that mean nothing, or mean much less than they appear to. A companion article covers the legitimate certification programmes, who runs them and what they verify.
“FDA registered” is the big one, and the rule sits in food law
The most common misleading claim in this category invokes the federal government. Bottles and websites carry phrases like “FDA registered,” “FDA approved facility,” “FDA certified” or a certificate with an official-looking registration number.
Here is the fact that matters, and getting its source right matters almost as much. Dietary supplements are legally foods, so the provision that governs a supplement facility’s registration is in the food regulations at 21 CFR 1.242. It says that assigning a registration number means the facility is registered with FDA, and then: “Assignment of a registration number does not in any way convey FDA’s approval or endorsement of a facility or its products.”
Registration is a notification. A firm tells FDA that a facility exists and what it does. No inspection is required to obtain it, no product is reviewed, and the number says nothing about quality.
Think about what that means on a shelf. Two bottles sit side by side. One is made in a plant that has never been inspected and has no history with the agency. The other is made in a plant FDA visited last year and left with a list of observations. Both plants are registered. Both can print a registration number. The number distinguishes nothing, which is precisely why it is attractive to put on a label.
Much of the strongest official language on this point is widely quoted and widely misattributed, so it is worth separating out. FDA’s blunt statements that registration is not approval, and that the agency issues no establishment registration certificates, appear on its medical device pages. In March 2021 the agency sent letters to 25 firms telling them to stop producing misleading “FDA registration certificates,” and that action concerned device establishments. The parallel drug regulation, 21 CFR 207.77, is sharper still: it says that any representation creating an impression of official approval because of registration or listing is misleading and constitutes misbranding.
Notice the asymmetry, because it is real and almost nobody reports it. The drug rule declares a registration-based approval claim to be misbranding in those words. The food rule, the one that covers supplements, says only that registration conveys no approval or endorsement. The regulatory text is harder on drugs than on foods. An “FDA registered” claim on a supplement is still actionable, but the route runs through general misbranding law and through the Federal Trade Commission’s authority over deceptive advertising rather than through the registration rule itself.
FDA’s own consumer update of 4 May 2026 puts the underlying position plainly: by statute, the agency does not approve dietary supplements for safety and effectiveness, and in fact many supplements can be marketed without FDA even being notified. The same page notes that the FDA logo is for official government use only and should not be used to misrepresent the agency. In its 2025 guidance for retailers and distributors, FDA lists misuse of its logo as one of the red flags that a product may be tainted.
Paying for certification is not the problem
A common and understandable reaction is that paying a certifier taints the result. The Federal Trade Commission has addressed that directly and come down the other way.
In the FTC’s guidance on certifications and seals of approval at 16 CFR 260.6, Example 8 describes a company that pays a reasonable fee to a nonprofit for a certification. The Commission’s analysis says: “Consumers would reasonably expect that marketers have to pay for certification.” Because the payment is expected, it is not a hidden material connection, and the claim is not deceptive on that basis.
So a fee does not make a seal fraudulent. Nearly every legitimate certification in every industry is paid for by the company being certified, because somebody has to fund the laboratory.
What does make a seal deceptive, according to the same guidance, is a short and specific list:
- The company awarded the seal to its own product without any independent evaluation.
- The standard behind the mark is controlled by the industry it certifies, so the claim of independence fails.
- The seal reflects membership in an organisation but is presented as though a product was evaluated.
- The seal implies an endorsement or certification by an independent third party that did not happen.
The guidance also says third-party certification does not relieve a marketer of the obligation to have evidence for every claim the certification reasonably communicates. That matters more than it sounds. A brand cannot point at a badge and stop. If the badge communicates that a product is pure, or effective, or safe, the brand still has to be able to back up whatever a reasonable consumer takes from it.
One caution on scope: the Green Guides are written for environmental marketing, and the specific rule about seals that fail to disclose their basis is framed in environmental terms, so this article takes only the general principle. Environmental and sustainability seals are covered elsewhere in this library.
The membership category deserves singling out, because it is the easiest to misread in good faith. Trade associations issue logos to members. A member logo is a statement that a company pays dues and agrees to the group’s conditions of membership. It is not a statement that anyone opened a bottle. Presented without that distinction, a membership mark can carry the same visual weight as a laboratory certification while standing for something entirely different.
Two cases that show what a hollow seal looks like
Enforcement records are more informative than principles, and two FTC matters show the business model with unusual clarity.
The first, filed in the Southern District of New York in June 2016, involved a company selling a trust seal to health and product websites. The certificate told visitors the site had been carefully evaluated by an independent medical doctor and found trustworthy and safe for purchases as of the review date. It carried a clinician-style photograph, a date stamp and a click-to-verify link.
The FTC’s complaint alleged that the operator had recruited two doctors, both family medicine specialists, and that they did not exercise medical expertise in granting the certifications. FTC staff writing about the case described a process that amounted to little beyond having doctors glance over a site. What the reviewers checked was whether a site stated that it had a return policy, displayed a phone number and mailing address, and claimed to use encryption. They did not verify that returns were honoured, that the phone was answered, that encryption was in place, or that any health claim on the site had evidence behind it.
The seal cost $39.95 a month after a free trial, or $99.95 for a first month followed by the same monthly fee. As of March 2015 it appeared on roughly 800 websites. The complaint also alleged the defendants falsely represented that the certification was issued by a nonprofit consumer protection organisation.
Sit with the economics for a moment. Two physicians, 800 sites, and a monthly subscription. There is no version of that arithmetic in which anyone read the evidence behind a health claim. The seal was not a failed review process. It was a product, and what the customer bought was the appearance of one.
The second case comes from a Federal Trade Commission staff blog post dated 12 January 2011, which described a proposed settlement with a certification programme that had told the public certified businesses were independently and objectively evaluated. The post reports a proposal rather than a final order, and gives neither a settlement date nor an outcome. According to the FTC, applicants were never asked about their actual practices, and buyers of the premium tier never received the site visits the programme advertised. What the programme required was payment: $189.95 for the basic certification and $549.95 for the premium one.
Two points from that case carry further than the facts. First, the company claimed endorsements from two industry associations whose owner was the same person who ran the certification business, and the failure to disclose that relationship was treated as a separate deception. Second, and more useful, the FTC charged that supplying businesses with the seal and the materials to advertise it was itself an unlawful deceptive act, on the legal theory known as means and instrumentalities. In other words, the seller of a worthless seal can be liable, not only the brand that displays it. Although that matter involved an environmental mark rather than a supplement, the three mechanics transfer directly: pay-only certification, self-dealing endorsements, and liability for the seal’s seller.
“Lab tested,” and the certificate nobody checked
Ask a supplement seller for proof and you will often be handed a certificate of analysis, usually shortened to COA. It is a laboratory document reporting what a sample contained.
The instinct is to worry about forgery. No documented case of a fabricated certificate of analysis for a dietary supplement was located for this article. Every COA falsification enforcement action found involved pharmaceutical ingredient manufacturers, which is a different industry and a different regulatory regime. So the forgery framing, however dramatic, is not where the evidence points.
The real mechanism is duller and much more common: a genuine certificate that nobody in the chain ever checked. A COA for a raw ingredient is typically produced by or for the supplier selling that ingredient. Under the manufacturing rules at 21 CFR 111.75, a supplement maker may rely on a supplier’s COA instead of testing an incoming component itself, but only if specific conditions are satisfied, including that the supplier has been qualified.
FDA documented exactly that failure in a warning letter to a Georgia contract manufacturer on 21 May 2020. The agency found the firm had chosen to rely on its suppliers’ certificates, and that it had never qualified a single supplier by confirming that those certificates were reliable. The paperwork existed. The verification behind the paperwork did not.
A warning letter dated 4 June 2026 to a different supplement firm shows the gap from the customer’s side, although the certificates themselves were not what FDA faulted there. When a customer asked that company for records, it supplied batch production records including certificates of analysis. What it could not supply were any records of complaint reviews and investigations, which FDA cited as a violation in its own right. The paperwork a buyer receives and the paperwork that would show whether anything went wrong are not the same documents. A certificate answers one narrow question about one sample. It is not a quality system, and it is not an audit.
“Lab tested” as a marketing phrase has no defined meaning in US regulation, no required scope, and no verifier. It does not say which laboratory, which batch, which analytes, or against what specification.
Four phrases, ranked by who verifies them
Four claims show up constantly. They are not equivalent, and the difference is who, if anyone, checks.
“GMP certified.” Current good manufacturing practice for supplements is a legal requirement under 21 CFR part 111, and FDA inspects facilities against it. But FDA issues no GMP certificate. A “GMP certified” mark therefore refers to a private audit by a third party, and its value depends entirely on which organisation performed it and against which standard. Manufacturing standards and the inspection record are covered in a separate article.
“Clinically proven.” This one has a real enforcer. The FTC can act against an unsubstantiated claim, and in April 2023 it sent notices to roughly 670 companies marketing over-the-counter drugs, homeopathic products, supplements and functional foods, warning of civil penalties. Those notices expressly covered misrepresenting that a product claim has been scientifically or clinically proven. Nobody screens the phrase before it is printed, so enforcement is after the fact. Claim substantiation and efficacy enforcement are treated in their own article.
“Lab tested.” No defined meaning, no required scope, no verifier. See the previous section.
“Doctor formulated.” No defined meaning and no verifier either. The 2016 FTC case shows what a doctor-review claim can amount to in practice: two family physicians checking whether a website said it had a phone number.
A plain ranking: one of these four has a legal standard behind it that FDA inspects against but does not certify, one has an enforcement body that acts after publication, and two have nothing at all.
What even a real seal cannot promise
The legitimate certification programmes do real work, and a reader should not conclude that all marks are worthless. They should conclude that a mark’s promise is narrow.
Two limits are worth holding onto. First, a certification measures a product against the label its own maker wrote. If that label declares a proprietary blend as a single total with no individual amounts, a passing result means the bottle holds those ingredients, in that order, adding up to that total. The programme has no power to demand a more informative label in the first place, which is the subject of a separate article in this library.
Second, even a mark that addresses contamination addresses it to a threshold, not to zero. The 2022 review of lead in mineral and multivitamin products in Annals of Pharmacotherapy makes the point concretely: the certification seals that attest to low lead levels still leave open the possibility that a child could receive more lead than FDA’s interim reference level. The same review found that without independent third-party testing, neither a pharmacist nor a consumer can tell a low-lead product from an outlier, which is an argument for certification rather than against it. Both things are true at once. Testing is the only way to know, and a passing result is a result against a specified limit, not a guarantee of absence.
A seal is a statement about a standard. Knowing the standard is the whole job.
Three questions, and when the seal is beside the point
A short test you can apply to any mark on any label or website:
- Who issued it, by name? If the mark does not identify an organisation you can look up, treat it as decoration. Invented marks are documented in the peer-reviewed literature, and they tend to use generic quality language rather than an issuer’s name.
- What exactly did they examine, and when? A mark should resolve to a standard, a scope and a date. Identity, purity, label accuracy and banned-substance screening are different tests. A mark that does not say which it performed is communicating a general quality claim, which is the hardest kind to substantiate.
- Can the claim be checked at the source? Legitimate programmes publish searchable lists of the products they certify. If a product’s own website is the only place the certification appears, that is the whole evidence base.
One situation makes the seal irrelevant. No mark of any kind speaks to whether a product works for a medical condition, and none substitutes for a diagnosis. If you are using a supplement for a symptom that is new, worsening, or persistent, or if you take prescription medicines, the thing to resolve is not the badge. Bring the actual container to a clinician or pharmacist so the ingredient list can be read directly.
And treat certain developments as reasons to be seen rather than to keep researching the seal. A heartbeat that races or skips, pain or pressure in the chest, fainting, severe stomach pain, or a yellow tint to the whites of the eyes accompanied by unusually dark urine all warrant prompt medical attention when they begin after a new supplement. No certification on the package changes that, because none of them were designed to.
The single most useful habit is also the simplest. Read the mark as a sentence rather than a symbol. Say out loud who issued it, what they tested and when. A legitimate certification survives that sentence easily. A decoration does not survive the first clause.
Sources
- Cornell Legal Information Institute. “21 CFR 1.242 – What does assignment of a registration number mean?.” law.cornell.edu
- Electronic Code of Federal Regulations. “21 CFR 207.77 – What legal status is conferred by registration and listing?.” ecfr.gov
- Electronic Code of Federal Regulations. “16 CFR 260.6 – Certifications and seals of approval.” ecfr.gov
- US Federal Trade Commission. “Complaint, FTC v. SmartClick Media LLC and Robert Vozdecky (S.D.N.Y., June 2016).” ftc.gov
- US Federal Trade Commission. “Lesley Fair, “Doctor who?” Business Blog, 21 June 2016.” ftc.gov
- US Federal Trade Commission. “Lesley Fair, “Let’s Make a Seal,” Business Blog, 12 January 2011.” ftc.gov
- US Food and Drug Administration. “Warning Letter: NutraCap Labs LLC, 21 May 2020.” fda.gov
- US Food and Drug Administration. “Warning Letter: Nutratech, LLC, 4 June 2026.” fda.gov
- US Food and Drug Administration. “Is It Really ‘FDA Approved’? Consumer Update, 4 May 2026.” fda.gov
- US Food and Drug Administration. “FDA Calls on Certain Firms to Stop Producing and Issuing Misleading ‘FDA Registration Certificates’ (3 March 2021).”
- Nutrients. “Crawford C, Lindsey AT, Avula B, Katragunta K, Khan IA, Deuster PA. “Label Accuracy and Quality of Select Weight-Loss Dietary Supplements Sold on or near US Military Bases.” Nutrients, 2024. DOI: 10.3390/nu16244369.” doi.org
- Annals of Pharmacotherapy. “White CM. “Lead in Mineral or Multivitamin-Multimineral Products.” Annals of Pharmacotherapy, 2022. PubMed. DOI: 10.1177/10600280211023328.” doi.org
- US Federal Trade Commission. “FTC Warns Almost 700 Marketing Companies That They Could Face Civil Penalties if They Can't Back Up Their Product Claims (13 April 2023).” ftc.gov
- Cornell Legal Information Institute. “21 CFR 111.75 – What must you do to determine whether specifications are met?.” law.cornell.edu
- Cornell Legal Information Institute. “21 CFR part 111 – Current Good Manufacturing Practice in Manufacturing, Packaging, Labeling, or Holding Operations for Dietary Supplements.” law.cornell.edu