Probiotic Supplement Testing: CFU Counts, Species Identification, and the Stability Data Most Brands Are Missing

The number on a probiotic label — 10 billion CFU, 50 billion CFU, even 200 billion CFU — tells you almost nothing about what’s actually alive in that capsule on the day a consumer takes it. We’ve processed enough probiotic submissions at our ISO 17025 accredited lab to recognize that this isn’t a fringe problem. It’s a structural one, baked into how the category approaches quality control.

The US probiotic supplement market crossed roughly $8 billion in retail sales in 2024 and is growing at approximately 8% annually, pushed by genuine consumer interest in gut health, immune support, and the expanding evidence base around the microbiome. That growth is outpacing the testing infrastructure supporting it. Brands that launched on the gut health wave are discovering that a single CFU count doesn’t satisfy an increasingly vigilant FDA, Amazon’s expanding compliance requirements, or quality-focused retail buyers who’ve seen enough recalls to know what a thin COA looks like.

Here’s what a thorough probiotic testing program actually requires — and where most brands are leaving gaps.

What “Billion CFU” Actually Means — and Doesn’t

CFU stands for colony-forming unit: a single viable bacterium (or clump of bacteria) capable of dividing and forming a visible colony on an agar plate under controlled conditions. The number printed on a probiotic label is typically determined at the time of manufacture, using culture-based plate counting under conditions specifically optimized for the organisms present.

That’s an inherently limited measurement. But the more important problem is what happens after the count is taken.

Probiotic organisms are biological materials. They respond to temperature, humidity, oxygen, light, and physical handling in ways that engineered chemical ingredients simply don’t. During a typical supply chain — from fill line to warehouse to distributor to retailer to consumer — a product can be exposed to thermal cycles, humidity fluctuations, and handling stresses that no accelerated stability model fully replicates. Peer-reviewed stability studies have documented that probiotics can lose 1 to 2 log CFU under suboptimal conditions. In practical terms: a product labeled at 10 billion CFU could reach a consumer with as few as 100 million viable organisms — a 99% reduction that is completely invisible on the package.

The labeling convention makes this worse. Under FDA’s 21 CFR Part 111, brands that state CFU “at time of manufacture” are technically compliant. But they’re setting a number that, for room-temperature products without robust formulation work, almost certainly won’t hold through a 24-month shelf life. Brands that state CFU “at time of expiration” are making the stronger and more consumer-meaningful claim — but it requires actual stability data to substantiate. Most brands don’t have that data when they launch.

Independent testing organizations, including ConsumerLab.com, have repeatedly found that a significant share of probiotic supplements sold at retail don’t meet their CFU label claims at time of purchase. This isn’t a statistical anomaly. It’s a predictable outcome of inadequate stability planning.

Species Identification: The Test Most Brands Treat as Optional

Here’s what surprises most people outside the analytical lab world: a standard plate count confirms that live bacteria are present. It doesn’t reliably tell you which ones.

Culture-based identification methods — growing colonies on selective media and examining morphological or biochemical characteristics — are adequate for genus-level confirmation. But genus is not strain, and strain is where the clinical evidence lives. Lactobacillus acidophilus, Lactobacillus rhamnosus GG, and Lactobacillus reuteri ATCC 55730 are all lactobacilli. They have fundamentally different clinical profiles, documented through strain-specific human trials. A brand that switches its probiotic raw material supplier and unknowingly receives a different Bifidobacterium longum subspecies may see a perfectly normal plate count and have no idea they’re delivering something different from what they formulated.

16S rRNA gene sequencing — specifically, amplifying and sequencing one or more hypervariable regions of the bacterial 16S ribosomal RNA gene — is the current standard for species-level confirmation. Whole-genome sequencing or multilocus sequence typing (MLST) goes further, enabling strain-level confirmation when a reference genome is available from the original strain provider.

For brands making structure-function claims supported by clinical literature on a specific strain, genomic confirmation isn’t a nice-to-have. It’s the only scientifically defensible way to show that the strain in the capsule is the strain the research actually studied.

Under 21 CFR Part 111.75, dietary supplement manufacturers are required to establish and verify the identity of incoming components. For probiotic raw materials, culture-based methods alone generally don’t satisfy that requirement at species resolution. A supplier COA listing “Lactobacillus acidophilus” with no sequencing data is, at minimum, a documentation gap — and in an FDA inspection, it’s a gap that gets noticed.

Stability Testing: The Data Most Probiotic Brands Are Missing

A point-in-time CFU count tells you what you had at the fill line. A stability program tells you what you’ll have at month 18, sitting in a distribution center in Phoenix in July.

The standard approach is real-time stability testing supported by accelerated runs following ICH Q1A(R2) guidance: 40°C / 75% relative humidity for 6 months as the accelerated condition, with 25°C / 60% RH for long-term data on room-temperature products. Refrigerated probiotics are tested at 5°C ± 3°C for long-term data. Both protocols require CFU enumeration at each time point — not just a physical description of the capsule.

What most probiotic stability programs miss is the interaction between the organisms and the dosage form itself. Enteric-coated capsules, delayed-release vegetable capsules, and sachet formats all create microenvironments that can either protect or accelerate die-off depending on formulation choices. We’ve seen products where the lyophilized probiotic blend looks excellent at time zero and degrades faster than expected at month 6 — not because of temperature excursions, but because moisture migration from the capsule shell into the fill was occurring under conditions the accelerated protocol didn’t fully stress-test.

Water activity (Aw) belongs in every probiotic stability protocol. Probiotic organisms in dried, lyophilized form are generally stable at Aw below 0.25. Products with Aw drifting toward 0.30 or higher tend to show significant CFU decline before the expiration date, particularly in the back half of shelf life. This number rarely appears on a probiotic COA, and its absence is telling.

What a Complete Probiotic Testing Panel Should Look Like

Based on submissions we’ve run for probiotic clients, here’s the core documentation any brand should be able to provide — built on validated methods appropriate for an ISO 17025 accredited dietary supplement testing lab:

CFU enumeration: Validated plate count with specified growth medium, incubation conditions (aerobic vs. anaerobic as appropriate — Bifidobacterium species require strict anaerobic conditions), incubation temperature, and duration. The method matters as much as the number.

Species and strain identification: 16S rRNA gene sequencing for species-level confirmation. Whole-genome or MLST data for strain-level claims tied to specific clinical literature.

Contaminant screening: Microbial limits per USP <2021> and <2022> for dietary supplements, with targeted testing for Salmonella spp., Staphylococcus aureus, E. coli, and Shigella spp. Products targeting children or immunocompromised populations warrant extended pathogen panels.

Heavy metals: Lead, cadmium, arsenic, and mercury per applicable limits — California Prop 65 action levels or USP dietary supplement limits, whichever is more stringent for the intended market. Probiotic raw materials grown on dairy, plant-based, or fermentation substrates can concentrate contaminants from growth media in ways that finished-product testing alone doesn’t always catch. Raw material testing matters here.

Stability data: CFU enumeration at time zero, 3 months, 6 months, 12 months, and 24 months (or through expiration, whichever comes first) under appropriate storage conditions, with documented water activity at each interval.

That’s five distinct test categories, each with its own validated method, defined acceptance criteria, and traceability requirements. A single-page COA with a CFU number and a basic microbial limits “pass” is not a probiotic quality program. It’s a starting point that leaves a brand exposed in ways most don’t discover until a retailer rejection or an FDA inspection surfaces the gaps.

The Retailer and Marketplace Pressure That’s Changing the Equation

The practical pressure driving brands toward more complete probiotic testing right now isn’t regulatory guidance — it’s the retail and marketplace channel. Amazon’s supplement compliance requirements increasingly specify third-party ISO 17025 accredited testing. For probiotics, that means a lab that has validated its enumeration methods for the specific organisms in the formula, not just one running a generic plate count.

Major natural product retailers are now requesting full stability data as part of supplier qualification. We’ve seen purchase orders contingent on 12-month CFU data through expiration — not time-of-manufacture data. That’s a reasonable ask from buyers who’ve had too many complaints about products that “didn’t work” and are tired of being caught between the consumer and the brand.

And here’s the thing about retroactive stability data: you can’t generate it after the fact on a timeline you don’t control. If you’re formulating a new probiotic SKU, build the stability program into your development schedule from day one. The data takes as long as it takes, and a buyer’s deadline won’t change that.

Before you release your next probiotic lot, make sure you can answer three questions clearly: What method was used to enumerate CFUs, and are the growth conditions documented? Has species identity been confirmed by sequencing, not just morphology? Does the formulation have stability data at or near the expiration date, not just at time of manufacture? If any of those answers is uncertain, that’s exactly where your testing program needs to start.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• Raw Material Identity Testing and COA Verification for Probiotic Ingredients — Ayah Labs covers incoming component qualification and identity verification for probiotic raw materials used in contract manufacturing.
• Health Canada NHP Evidence Requirements for Probiotic Products Sold in Canada — Androxa breaks down what the Natural Health Products Directorate requires for probiotic NHP license applications in the Canadian market.
• FDA Regulatory Strategy for Dietary Supplement Structure-Function Claims — Aurora TIC walks through the regulatory pathway for substantiating probiotic health claims under DSHEA, including the notification and disclaimer requirements.