Reading a probiotic label: strain identifiers and CFU at end of shelf life

The probiotic category is one of the most diluted in supplement retail. Tens of thousands of products on the shelf, hundreds of bacterial species, and almost no consistency in how labels disclose what is actually in the capsule. Two pieces of information determine whether a probiotic is worth buying for a specific purpose: the precise strain identifier and the colony-forming unit count guaranteed at the end of shelf life. Most labels disclose neither cleanly.

Genus, species, and strain are three different levels of specificity

"Lactobacillus" is a genus. "Lactobacillus rhamnosus" is a species. "Lactobacillus rhamnosus GG" is a strain — and the strain is the only level at which clinical evidence applies. The International Scientific Association for Probiotics and Prebiotics (ISAPP) consensus definition explicitly requires strain-level identification for any health claim (PMID: 24912386).¹ Two products both labelled "Lactobacillus rhamnosus" can contain entirely different organisms with entirely different effects. A label that does not include a three- or four-letter strain designation (LGG, DSM 17938, NCC2818, Bb-12) is not giving you the information needed to map onto trial evidence.

The CFU count problem

Most probiotic labels list a colony-forming unit count "at the time of manufacture" — typically 10 billion, 25 billion, or 50 billion CFU per capsule. Probiotic organisms are alive and die during storage. A 2010 quality audit of commercial probiotic products found that 33% contained fewer than half the labelled CFU count at the time of consumer purchase (PMID: 20336366).² A 2022 European market survey of 50 probiotic products found a median 27% gap between labelled and measured CFU at point of sale (PMID: 35462457).³ The number that matters is CFU at end of shelf life — what the manufacturer guarantees will still be alive when the consumer takes the capsule on the expiration date. Reputable manufacturers state this explicitly.

Matching strain to trial: a few high-confidence examples

For pediatric acute infectious diarrhoea, Lactobacillus rhamnosus GG (LGG) and Saccharomyces boulardii CNCM I-745 have the strongest trial base, reducing diarrhoea duration by roughly one day in pooled analyses (PMID: 32077460).⁴ For antibiotic-associated diarrhoea in adults, the same two strains plus Lactobacillus reuteri DSM 17938 have replicated benefit (PMID: 28192103).⁵ For irritable bowel syndrome, the multi-strain VSL#3 / Visbiome formulation has the most consistent evidence at trial-grade dose (PMID: 24834008).⁶ The point of these examples is that a product labelled "8 strain blend, 50 billion CFU" without specific strain identifiers cannot be matched to any of this evidence.

The "shelf-stable" versus refrigerated debate

Most commercial probiotics are now sold as shelf-stable products, typically using lyophilised (freeze-dried) bacteria in moisture-resistant blister packs. A 2019 storage study confirmed that properly lyophilised Lactobacillus and Bifidobacterium strains in low-moisture packaging retain viability comparable to refrigerated products for up to 24 months (PMID: 31262987).⁷ Refrigeration is no longer a quality marker on its own — what matters is the manufacturer's stability dossier and the end-of-shelf-life claim, both of which the better manufacturers publish.

Spore-formers, prebiotic blends, and "10x more potent" claims

Bacillus coagulans and other spore-forming probiotics are inherently more stable than lactobacilli and bifidobacteria — their endospore form survives stomach acid and storage without enteric coating. They are a reasonable choice for products that cannot be guaranteed cold-chain, but they should not be assumed equivalent in clinical effect to the more-studied lactobacilli. Marketing claims like "10 billion CFU equivalent to 100 billion of competing strains" are not based on any standardised potency measure. Synbiotic blends combining probiotics with prebiotic fibres can have additive effects, but the dose of the prebiotic component matters — under 3 g/day of inulin or FOS is typically below the threshold seen in microbiome modulation trials (PMID: 30853595).⁸

The checklist for buying a probiotic in 2026

Before adding a probiotic to the basket, the label should disclose strain-level identifiers for every organism in the formula, a CFU guarantee at expiry rather than manufacture, a quality manufacturing certification (NSF, USP, GMP), and a published stability dossier or third-party verification of organism viability. The pre-purchase question is whether the strain-and-dose combination on the label matches an indication the consumer has and an indication with clinical evidence behind it. If it doesn't, the product is generic gut bacteria of unknown identity at an unknown surviving dose — a category in which most of the probiotic market still operates.