When the Bacteria on the Label Aren’t the Bacteria in the Bottle
Independent testing of commercial probiotic products has repeatedly found that what’s on the label and what’s in the capsule are different things
Wrong strains. Missing strains. Organisms present in quantities a fraction of what was promised.
In one documented case: no viable organisms at all.
There is no mandatory pre-market testing requirement that would catch any of it.
Here is a straightforward question that should have a straightforward answer: does the probiotic supplement you are taking contain what the label says it contains?
The answer, based on a body of independent testing research spanning more than a decade across multiple countries, is: not reliably, and sometimes not at all.
This is not a fringe finding from a single discredited study. It is a recurring, replicated observation from peer-reviewed research, independent laboratory testing, and legal proceedings that have resulted in multi-million dollar verdicts.
The probiotic supplement industry is selling products with label claims that, in a meaningful proportion of cases, do not reflect the product’s actual contents – and is doing so in a regulatory environment that requires no independent pre-market verification that they do.
What the Research Has Found
A study evaluating 52 commercial probiotic dietary supplements found that 17 samples – 33 percent – were below label claim for CFU count prior to their expiration dates.
Thirty-three percent. One in three products, tested before expiry, containing fewer organisms than the label promised.
A separate assessment found that if an adequate label is defined as one containing specific and valid bacterial names, with no spelling errors and a clear statement of the number of viable organisms expected, then only eight of 25 products were properly labelled.
If a more stringent definition is applied, requiring the product to identify the specific bacterial strain present, then no products were adequately labelled. One product claimed to contain Lactobacillus acidophilus and had no viable growth on any replicate.
Testing was repeated using different capsules from the same bottle and the absence of growth was confirmed. A capsule sold as a living organism product, confirmed by repeated testing to contain no living organisms.
A UK study of commercial probiotics evaluated bacterial viability and label accuracy. The only UK study previously identified had evaluated seven human products in 2016, and fewer than half of those met their declared culture concentrations.
The 2025 study found that human products delivered greater variability in both species composition and stated CFU count.
A Polish study found significant discrepancies in the quality of probiotic supplements available on the market. Only a few products maintained bacterial counts consistent with label claims, while most contained considerably low viable cells.
In several cases, undeclared probiotic microorganisms were identified – organisms not listed on the label were found inside the capsule. Not just fewer of what was promised – but the presence of organisms that were never disclosed.
The capsule contained bacteria the label didn’t mention. In a product category where specific strain identity is the entire basis of any clinical claim, this is not a minor quality issue. It is a fundamental failure of product integrity.
The Legal Record
The mismatch between label claims and actual contents has moved, in recent years, from academic papers into courtrooms – and the findings there are at least as striking as the laboratory research.
In the Zesty Paws class action, independent testing of their Aller-Immune Bites found 72,000 CFU of lactobacillus per serving instead of the promised 250 million.
Not a modest shortfall. Not a 20 percent discrepancy attributable to normal manufacturing variation. Less than one-thousandth of a percent of the stated amount.
In a separate case involving VSL#3 – a high-dose multi-strain probiotic with genuine clinical evidence behind it – a manufacturer changed the formulation while continuing to market the product using clinical data generated with the original formula.
A Maryland jury found this was false advertising under the Lanham Act and awarded over $18 million. The appeals court upheld the verdict. In April 2025, a federal judge certified 10 classes of consumers in the related RICO case.
The original VSL#3 had clinical evidence. The reformulated product did not – but was being sold as though it did. The jury found this fraudulent. The awards are substantial.
And the case established something important as legal precedent: you cannot change a probiotic formulation and continue claiming the same clinical evidence.
Why This Keeps Happening
The strain fraud problem has three structural causes, each reinforcing the others.
No mandatory pre-market testing. In the UK, probiotic supplements are regulated as food supplements. The manufacturer is responsible for ensuring their product meets its label claims. There is no requirement for independent laboratory verification before the product reaches market.
The MHRA does not test probiotic supplements for label accuracy before sale. The FSA has no systematic pre-market testing programme for supplement CFU counts. The manufacturer can label their product with any CFU count and strain composition they choose, and no regulatory body will verify this claim before the product is on the shelf and the consumer has paid for it.
Post-market enforcement exists in principle. In practice, it is resource-limited, complaint-driven, and sufficiently infrequent that a manufacturer calculating the expected cost of non-compliance – the probability of being tested multiplied by the likely consequence – would find the expected value of honest labelling difficult to distinguish from the expected value of dishonest labelling.
The measurement problem. Standard CFU counting is imprecise. The test typically sees the species, not the exact strain. If a product has multiple strains, the test often ends up giving one total number instead of a breakdown for each strain.
Some cells won’t replicate or simply won’t grow under standard lab conditions, meaning the tally misses them entirely.
This creates a technical grey area that manufacturers can exploit. A product that claims “10 billion CFU of Lactobacillus acidophilus NCFM” can be tested and found to contain 10 billion CFU of something in the Lactobacillus acidophilus species – without those organisms necessarily being the NCFM strain.
Standard testing cannot distinguish at the strain level. More sophisticated metagenomic testing can – but it is expensive, is not routinely applied, and is not required.
The shelf-life attrition problem. Probiotic organisms die over time. A product that genuinely contained 20 billion CFU at the point of manufacture may contain 4 billion by month six and fewer by month twelve.
This is not fraud – it is biology. But it becomes fraud when the label claims 20 billion with no qualification about when that count was measured, no end-of-shelf-life guarantee, and no manufacturing overage built in to account for inevitable decline.
The industry standard – largely honoured in the breach – is to manufacture with a significant overage: to put in 40 billion if the label claims 20 billion, accounting for the expected attrition over shelf life.
Some manufacturers do this. Many do not. The consumer has no way to know which category their product falls into without independent testing.
The Strain Specificity Problem: Why It Matters More Than You Think
Even where viability testing shows adequate CFU counts, strain misidentification creates a problem that CFU testing alone cannot detect.
The clinical evidence for probiotics is strain-specific. Lactobacillus rhamnosus GG has clinical evidence for antibiotic-associated diarrhoea prevention. Lactobacillus rhamnosus HN001 has clinical evidence for eczema reduction in infants.
These are the same species. They are different strains. Their clinical evidence is not interchangeable.
Standard CFU testing cannot distinguish between Lactobacillus rhamnosus GG and Lactobacillus rhamnosus HN001. It sees the species. It does not see the strain. A product that labels itself as containing LGG may actually contain a different Lactobacillus rhamnosus strain – one with different biological properties and a different clinical evidence base – and standard testing would not catch this.
The regulatory requirement for strain-level disclosure on probiotic labels is, in the UK, essentially non-existent. A label that says “Lactobacillus rhamnosus” without specifying the strain designation is entirely compliant with current labelling requirements.
The clinical study you may have read about – the one that led you to seek out a product containing this organism – was almost certainly conducted on a specific strain at a specific dose. The product you purchased may contain a different strain at a different dose, without any requirement to tell you this.
What the Industry Does About This
The honest answer is: not enough, and not systematically.
Some manufacturers invest in third-party certification – NSF International, Informed Sport, or the US Pharmacopeia’s verification programme – which requires independent laboratory testing of label claims before certification is granted.
This is genuinely meaningful. Products carrying these certifications have been independently verified to contain what their labels say.
The majority of mass-market probiotic products carry no such certification. Certification is voluntary. It costs money. It exposes the manufacturer to the risk of failing the test. The commercial incentive to seek voluntary certification is weaker than the commercial incentive to avoid the risk of discovery.
Trade associations have developed voluntary guidelines. The International Probiotics Association publishes standards for label accuracy and strain disclosure. Responsible manufacturers adhere to them. Irresponsible ones don’t, and face no mandatory consequence for the difference.
The regulatory agencies – MHRA, FSA in the UK; FDA, FTC in the US – have the authority to take action against products with demonstrably false label claims. They exercise this authority sporadically, in response to complaints or egregious cases, rather than through systematic market surveillance.
The gap between the rate at which label inaccuracy occurs and the rate at which it is acted upon is large, and the industry operates comfortably within it.
What You Should Actually Do
The only reliable protection against strain fraud is independent verification. This means:
Buying products that carry NSF International, USP Verified, or Informed Sport certification – the three most credible third-party verification marks available in the UK and US markets. These are not infallible, but they represent genuine independent testing that the uncertified majority of the market has not submitted to.
Seeking out manufacturers who publish their third-party Certificate of Analysis – the laboratory testing document that records the actual CFU counts and organism identities found in the finished product.
This document should be specific to the batch you’re buying, not generic to the product line. Its existence and accessibility is itself a signal of manufacturing discipline.
Being sceptical of any product that lists strain names without strain designations. “Lactobacillus acidophilus” is a species. “LA-5,” “NCFM,” or “DDS-1” are strains. The strain designation is the clinically meaningful unit of information.
Its absence from the label is a red flag.
Understanding that the CFU count on the label was measured at manufacture – not at the time of consumption – and that no product without an explicit end-of-shelf-life guarantee can reliably tell you how many organisms you are actually taking.
The probiotic supplement you are paying for may contain exactly what it says. It may contain less. It may contain different strains. It may contain no viable organisms at all.
Without independent verification, you are relying on the honesty and manufacturing competence of an industry operating in a regulatory environment that has never systematically required it to demonstrate either.
That is the market you are shopping in. You deserve to know that before you hand over your money.
See more here substack.com
Header image: Consumer Lab