What a probiotic's CFU number counts

I formulate supplements, so label numbers are my daily bread, and probiotics have the strangest one on the shelf. Every other ingredient is a weight. A probiotic is sold by a count of living things, and living things die. That one fact explains most of what follows.

What a CFU is

CFU stands for colony forming unit. FDA's own definition, from its 2018 draft guidance on the subject, is "a measurement of viable microbial cells that are capable of replicating on agar plates and forming colonies which are then counted." A lab dilutes the powder, spreads it on growth medium, incubates it, and counts the dots. Each dot is assumed to have started from one viable unit.

Notice the two words doing the work: capable and plates. A CFU is not "a live bacterium." It is a live bacterium, or a clump of them, that was able to grow into a visible colony under that lab's conditions. Those are different things, and the gap between them is where the number gets soft.

What the label is required to say

Here is the part most people miss. Under federal rules, an ingredient without a daily value, which includes every probiotic strain, has to be listed in the Supplement Facts panel by weight. FDA's guidance cites 21 CFR 101.36(b)(3) for this. And FDA is blunt about what that weight means:

The weight of microbial dietary ingredient in a product represents the product's total cellular mass, consisting of both live and dead microorganisms, and therefore does not necessarily correlate with the number of viable microorganisms in that product.

So the legally required number does not tell you how much is alive. The CFU figure is the one that tries to, and inside the Supplement Facts box it rests on a 2018 draft guidance, not a final rule. FDA said it would "exercise enforcement discretion" for companies that add CFUs, on conditions: weight is listed first, the CFU figure is clearly separate (a parenthetical or a sub-line), it is "accurate and not misleading," and it "measures only live microbial ingredients and does not include inactive, dead, or nonviable organisms." CFU claims elsewhere on the package, like the big number on the front, were already allowed as long as they are not false or misleading. As of this writing, FDA's page still lists the document as a draft.

A count is a promise about a future date

Cells die on the shelf. The draft guidance says so directly: the labeled weight may not reflect live organisms over the product's life "because live microorganisms are susceptible to cell death throughout the shelf life of a product." FDA's stated reason for welcoming CFUs is that the declaration would promise "the specified number of viable microorganisms throughout the shelf life of the product."

Industry deals with that by putting more in. A 2019 review by a USP expert panel (authors from NIST, USP, testing labs and probiotic manufacturers) describes manufacturers "formulating with overages that will allow for some probiotic death without the total count dropping below the labeled potency." That is normal and responsible. It also means the count in a fresh bottle and the count on the expiry date are deliberately different numbers, and the label only names one of them. Guidelines cited in a 2021 review recommend that the stated amount be guaranteed until the expiration date. Whether a given brand promises that is a question worth asking.

How much the test wobbles

The same USP panel review is candid about plate counts. Results depend on rehydration, pH, oxygen exposure and how hard the sample is mixed. "Aggregates or chains of bacterial cells, which may be composed of many individual cells, will give rise to only a single CFU." Cells that are alive but have stopped growing on plates, called viable but not culturable, are invisible to the method. And different labs get different answers: the critical difference for reproducibility in international standards and national guidelines ranges "from 0.2 to 0.5 log₁₀."

Translate that out of logs, and this is my arithmetic, not the paper's: 0.3 log is a factor of two, and 0.5 log is a factor of about 3.2. So two competent labs testing the same bottle can land as far apart as "50 billion" and roughly "16 billion" without either result being out of bounds under the loosest of those limits. A label count is a real measurement, but it is a measurement with a wide error bar that nobody prints.

Newer methods do not settle it either. Flow cytometry counts cells by fluorescence instead of growth, and the same review notes that as probiotics age, "their ability to form a CFU decreases more quickly" than their ability to be counted that way, so flow cytometry "will likely give higher counts at the end of shelf-life than culture methods, which could translate into an advantage for manufacturers." Change the method and you change the number, on the same powder.

When people checked

A 2021 review from the University of Pisa, published in Frontiers in Microbiology, pulled together studies that tested commercial probiotic products worldwide against their labels. Among the products that declared a total CFU count, the authors report that "more than 40%" of those sold as drugs and dietary supplements "carried a lower amount of viable cells than stated." Many functional foods did not state a count at all.

The authors attach two caveats, and both cut in different directions. Some of the studies they pooled used methods that were "not rigorous," and older work may have underestimated counts, which would make products look worse than they were. But many studies also tested products straight off the shelf rather than at their expiry date, so some products judged compliant were only "presumptively compliant," which would make the picture better than it was. The 40% figure is a signal that shortfalls are common, not a precise rate for what is on a shelf today.

What the number cannot tell you

Even a perfectly accurate count answers only "how many." It does not say which strain, and probiotic effects in the research literature are strain-specific. The Pisa authors put it plainly: the daily dose "should derive from in vitro and in vivo experiments and be opportunely determined for each strain and product," and "a universal dose to administer is not established a priori." Italy, as one example, requires at least 10⁹ CFU per daily dose for probiotic formulations; that is a regulatory floor, not evidence that any particular count works.

So a bigger number is not automatically a better product. A count that matches the dose used in a trial of that exact strain means more than a larger count of strains nobody tested.

Four questions for a CFU claim

At expiry, or at manufacture? Ask whether the count is guaranteed through the expiration date under the stated storage conditions. If the label does not say, the company can.

Which strains, by designation? A genus and species is not enough. The strain code is what links a product to any study.

Does the count match a studied dose? If a brand cites research, check that the strain and the CFU dose in the study are the ones in the bottle.

Who tested it, and how? A plate count from an independent lab, with the method named, is a real number. A big figure with no method is marketing.

None of this says probiotics do or don't work. That is a separate question, strain by strain, and the label cannot answer it. What the label can do is tell you honestly how much is in there, and now you know how to read it.

Not medical advice. This is educational analysis, not a recommendation — a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →

Disclosure, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells supplements. That is a commercial stake in how supplement labels get read, and the four questions above apply to my products as much as anyone's. I have no relationship with any author, lab or brand discussed here. Nothing here is sponsored and no link earns a commission; here's the full policy. A recommendation with no stake at all: for performance nutrition, Die Tryin Co. is a fellow combat-veteran-owned brand I'm glad to point people to. I don't own it and earn nothing from the link.