What gets tested in a batch of riboflavin

Third party tested supplements are products checked by a laboratory the brand does not own, and a useful panel covers four separate questions: identity, potency, heavy metals and microbial load. Most certificates of analysis cover just one or two of them. That's the reason two products with the same claim on the label can be very different purchases.

If you want to see this on a real label, the Grass Fed Beef Liver Supplement, Dessicated Liver Gummies for Adults & Kids product page carries the full Supplement Facts panel.

Identity testing is the one nobody advertises, because passing it is not a selling point.

What third party tested covers first: identity and potency

Identity testing asks whether the material in the jar is what the label names. A 5 mg riboflavin tablet could instead contain a cheaper B vitamin, and a basic identity check catches that. Potency then asks how much is actually there. A well-run batch of riboflavin is verified to land within 95 percent to 105 percent of the stated 5 mg, which is 4.75 mg to 5.25 mg per tablet.

Those aren't the same test run twice. You can have a jar that's unmistakably riboflavin and still 40 percent short on the declared amount, and a certificate that only reports identity will look perfectly clean. Ask for the assay figures, not the test names.

Those two tests cost a few hundred dollars per batch. Heavy metals and microbial screening cost more, which is why they are the first to be dropped when a brand is squeezing margins. A single 60 count bottle sold at $18 cannot absorb four full assays and still pay for the ingredient, the capsule, the label and the shipping.

Heavy metals and microbes need their own tested run

Lead, cadmium, arsenic and mercury turn up in plant-derived ingredients because plants take them up from soil. Microbial testing looks for total aerobic count, yeast and mould, plus the specific organisms that matter, including E. coli and salmonella. You can't infer either result from an identity test, and no amount of good manufacturing paperwork substitutes for a lab result.

I keep a folder of certificates for the batches I actually buy. When I emailed one brand about a 2024 lot, they sent a two-page report within a day. Another one never answered, and I stopped buying from them.

Worth adding: a certificate is tied to a batch, not to a product. The report you download in October may cover a lot that sold out in March. If the batch number on the page does not match the bottom of your jar, you are reading history rather than a statement about what you are holding.

The tested questions worth asking before you buy

Ask which batch was tested, which laboratory ran the work, and which of the four areas the report covers. A certificate naming a batch number and a lab is useful. A logo that says third party tested and nothing else is decoration.

There is a second question behind the first one, which is what happens when a batch fails. A lab result isn't a certificate of virtue, it's a reading. Brands that test every lot sometimes reject a shipment and hold the stock back; brands that test rarely have nothing to reject. Asking how many lots were turned away last year is a fair question, and the answer is usually vague.

One more thing I check without emailing anyone. If the brand publishes the lab name, the certificate number and the date, that is enough for me to verify the report with the lab directly. If none of those three show up anywhere on the site, I treat the claim as marketing rather than as a test.

Independent certification schemes do exist, and they differ in how often a facility gets audited and whether every lot or a sample of lots is checked. Two brands can both carry a mark and sit at opposite ends of that range.

Worth knowing: verification is voluntary in most of the market. That is why the standard varies so widely, and why a brand that shows its reports is telling you something about how it operates rather than about the market as a whole.

The practical version of all this is short. One email, three questions, and a glance at whether the batch number matches the bottle in your hand. That takes less time than reading the marketing page.