Probiotic potency is a moving target: a batch that tests at 50 billion CFU per capsule on release may test well below the label claim twelve months later. The buyer's problem is that the label claim is a legal commitment at end of shelf life, while the certificate of analysis usually describes the product on the day it was made. Closing that gap requires deliberate overage, stability data and a label that matches both.
What CFU means on a label and in a specification
CFU stands for colony-forming units, a count of viable microorganisms able to grow on a specified medium under specified conditions. Two products labelled "10 billion CFU" can differ substantially if one guarantees the count at release and the other guarantees it through expiry. Under US labelling rules, the Supplement Facts panel must state the quantity of a dietary ingredient per serving (21 CFR 101.36), and that quantity must remain accurate through the product's shelf life. The FDA does not publish a separate probiotic potency regulation, so the enforceable standard is the general requirement that labels be truthful and not misleading, combined with the CFR 111 requirement that finished-product specifications be met for the duration of shelf life.
In practice, a defensible specification states a release target and an expiry target. Example: release not less than 30 billion CFU per capsule, expiry not less than 10 billion CFU per capsule, with the label claim set at 10 billion. The overage between release and label is the manufacturer's margin against viability loss. Typical industry overage at release ranges from about 1.5x to 3x the label claim, depending on strain, dosage form and shelf life; treat that as a planning range, not a rule.
Why overage is not simply "add more powder"
Overfill has cost, capsule-size and sensory consequences. A 10 billion label claim with 3x overage means 30 billion CFU of viable cells at release, plus the non-viable biomass that comes with the culture. At high doses this can push a size 0 capsule toward size 00 or force a two-capsule serving, which changes the label, the packaging and the consumer experience. Overfill also raises the risk that the finished product exceeds a specification ceiling if one exists, and it increases exposure to water activity and oxygen during encapsulation.
The more reliable levers are strain selection, dosage form and packaging. Well-documented commercial strains differ in oxygen and moisture tolerance; spore-forming species such as Bacillus are generally more robust than many Lactobacillus and Bifidobacterium strains. Blister packaging or induction-sealed HDPE bottles with desiccant outperform simple bottles in humid conditions. If the product is a capsule, the choice between vegetarian and gelatin shells, and the use of a moisture barrier, affects viability. Buyers sourcing capsule manufacturing should ask specifically how the shell and packaging were selected against stability data.
Stability testing: timelines, conditions and what the data must show
Stability is where probiotic claims are won or lost. ICH Q1A(R2) sets out the general framework: long-term, intermediate and accelerated conditions, with testing frequency typically every three months in year one, every six months in year two, and annually thereafter. For dietary supplements in the US, 21 CFR 111 requires that the shelf life be supported by data, and the manufacturer must have a written stability program. Accelerated conditions are useful for screening but are a poor predictor for live cultures, because viability loss is often driven by water activity and oxygen rather than temperature alone. A realistic program includes real-time data at the intended storage condition, typically 25°C/60% RH or 30°C/65% RH, and often an additional 30°C/75% RH arm for products distributed in warm climates.
Ask for the actual assay method, not just the result. USP <711> covers microbial enumeration and is a common reference for total viable count, but it does not by itself validate a strain-specific claim. If the label names a strain, the method should confirm that strain, for example by selective media or genetic identification. Water activity is a leading indicator of viability loss and should be measured on the finished product, not only on the raw material. A stability protocol that reports only "total lactic acid bacteria" is weaker than one that reports the labelled strains separately.
Label claims that survive scrutiny
Three label decisions drive most disputes. First, whether the CFU figure is stated "at time of manufacture" or "through expiry". The first is easier to support and weaker commercially; the second is what most buyers want and requires stability data to the end of shelf life. Second, whether strains are named. Naming a genus and species is a lower evidentiary bar than naming a specific proprietary strain, which implies the strain itself was tested. Third, whether the product carries structure/function claims. In the US these must be accompanied by the standard disclaimer and must not be disease claims; the manufacturer's role is to ensure the claim text matches the substantiation file the brand owner holds.
Ingredient declaration also matters. 21 CFR 101.4 governs ingredient listing, and 21 CFR 101.36 governs the Supplement Facts panel. Cultures may be declared as a dietary ingredient with the CFU amount, and excipients such as microcrystalline cellulose, magnesium stearate or cryoprotectants must be listed. In the EU, Regulation (EU) No 1169/2011 governs food information, and probiotic claims fall under the separate health claims regime; a "supports digestive health" claim needs an authorised claim or must be avoided. A contract manufacturer should not be the source of your claim wording, but it should flag when the wording on your artwork is inconsistent with the formula or the evidence file.
Documents to request before you sign a purchase order
- Strain identity and deposit documentation for each culture, including the culture collection number where a proprietary strain is named.
- Release and expiry specifications stating CFU per serving at both points, plus water activity and appearance limits.
- Stability protocol and raw data covering the actual dosage form and packaging, with real-time points, not only accelerated.
- Assay method validation for the CFU method, including strain confirmation where strains are named.
- Certificate of analysis for the finished batch, showing the release result against specification and the test method used.
- Allergen and excipient declaration matching the label, including carrier media used in culture preparation.
- Heavy metals and microbial safety data referenced to USP <232>/<233> for elemental impurities where applicable, and to the manufacturer's own safety specifications.
- Facility registration and inspection status, and evidence of a current GMP audit under 21 CFR 111.
- Change control notice covering culture supplier changes, carrier changes and packaging changes.
For products imported into the US, the brand owner is the importer and carries FSMA Foreign Supplier Verification Program obligations under 21 CFR 117, and must file Prior Notice under 21 CFR 1005. Confirm who holds those responsibilities in your supply chain before the first shipment.
Failure modes and how they show up
The most common failure is a release result that passes and an expiry result that does not, usually because overage was set from a single accelerated timepoint. The second is a strain substitution by the culture supplier without change notification, which invalidates the identity data behind a named-strain label. The third is moisture ingress through a packaging format that was never stability-tested, often a bottle with a desiccant that was validated in a different size. The fourth is a label that states a CFU count "at time of manufacture" while marketing copy implies through-expiry potency. The fifth is an excipient change that alters water activity and accelerates viability loss. Each of these is preventable with a written specification, a stability protocol and a change-control clause.
Questions to ask your supplier
- Is the label claim guaranteed at release or through expiry, and what real-time data supports the expiry figure?
- What overage do you apply at release, and how was it derived for this strain and dosage form?
- Which strains are named on the label, and what identity method confirms each one in the finished product?
- What water activity limit do you set, and do you test it on the finished capsule or only on raw material?
- Has this exact packaging format been through stability testing, including the desiccant or blister configuration?
- What is your change-control process for culture supplier, carrier and packaging changes, and how much notice do I get?
- Can I see a recent third-party GMP audit report and your facility registration status?
- Who is responsible for FSVP and Prior Notice if the product is imported?
If you are comparing suppliers on price alone, ask for the stability package first. A lower price per kilogram often reflects a lower release overage, which shifts the risk of a failed expiry claim onto the brand owner.
Frequently asked questions
What is a normal CFU overage for a probiotic supplement?
Typical industry overage at release is roughly 1.5x to 3x the label claim, depending on strain, dosage form, packaging and target shelf life. The number should be derived from real-time stability data for that specific product, not copied from another SKU. A supplier that cannot show the derivation is asking you to accept the risk.
Can a label state CFU "at time of manufacture"?
It can, but the wording must be clear and the marketing must not imply through-expiry potency. Most buyers prefer a through-expiry claim because it is easier to defend and more useful to consumers. Whichever you choose, the Supplement Facts quantity must be accurate for the period the label describes, and the shelf life must be supported by stability data under 21 CFR 111.
How long does probiotic stability testing take?
Real-time data to a two-year expiry takes two years. Accelerated arms can give early signals in three to six months, but they do not replace real-time data for live cultures. Plan for a bridging study on the first commercial batch and budget lead time accordingly; a new SKU with no stability history should not be launched on accelerated data alone.
Do probiotic capsules need a different stability protocol than powders?
Yes. Capsule shells add a moisture barrier and a dissolution step, and the shell material affects water activity inside the capsule. The protocol should test the finished capsule in its commercial packaging, including any desiccant, at the intended storage conditions. Powder in a sachet and powder in a bottle are different products for stability purposes.
What certifications should a probiotic contract manufacturer hold?
At minimum, evidence of GMP compliance under 21 CFR 111 for the US market, and for the EU, compliance with the applicable food hygiene and, where relevant, GMP expectations. Third-party audits and scheme certifications are useful signals but do not replace the batch-level documents. Review the certifications a supplier holds against the markets you actually sell into, and confirm the scope covers probiotics rather than only general powders.
Probiotic manufacturing rewards buyers who specify the expiry claim first and work backwards. Set the label claim, require real-time stability data for the exact dosage form and packaging, agree the overage in writing, and make change control a condition of the purchase order. The supplier's job is to meet the specification; the brand owner's job is to make sure the specification is one the label can defend.




