Probiotics Dosage: How Many Billion CFU (and Why Strain Matters More)
Informational summary of published research — not medical advice. Probiotics are not for everyone: if you are immunocompromised, critically ill, or have a central line, talk to your doctor first.
Quick answer
Most effective doses are 1–10 billion CFU/day (some multi-strain products use 50–100+ billion). But CFU is the wrong thing to shop on — the effect is strain-specific. 10 billion CFU of the strain proven for your goal beats 100 billion of an untested one. Match the strain first, then the CFU is whatever its trials used.
As an Amazon Associate we earn from qualifying purchases. Picks are matched to the strain studied for each goal, never commissions.
Strain beats CFU — this is the whole game. "Probiotic" isn't one ingredient; it's hundreds of distinct strains, and what one strain does says nothing about another, even in the same species. The strain code — the letters/numbers after the species, like Lactobacillus rhamnosus GG or Bifidobacterium infantis 35624 — is the part of the label that maps to a real trial. A "50 billion CFU" tub listing only genus names ("Lactobacillus, Bifidobacterium") with no strain codes and no condition is a marketing number, not a clinical one.
Match a strain to your goal
Pick what you want it for — you'll get the strain and CFU dose that actually has evidence.
Probiotic dose by goal (CFU & strain)
| Goal | Best-studied strain | CFU dose | Evidence |
|---|---|---|---|
| Antibiotic-associated diarrhea | S. boulardii or L. rhamnosus GG | 5–20 billion/day | Strongest |
| Acute infectious diarrhea | S. boulardii / LGG | ~10 billion/day | Good |
| C. difficile prevention (on antibiotics) | Multi-strain | Higher — with your doctor | Good (higher-risk patients) |
| IBS symptoms | Strain-specific (e.g., B. infantis 35624) | 1–10 billion/day | Mixed |
| General "gut health" | No specific strain proven | 1–5 billion/day | Weakest |
CFU is a secondary detail — match the strain to your goal first (use the matcher above), then take the CFU its trials used.
What actually has evidence (by goal)
- Antibiotic-associated diarrhea — strongest. Probiotics roughly halved the risk; S. boulardii (18.7%→8.5%, Szajewska 2015) and L. rhamnosus GG (Szajewska & Kołodziej 2015) are the best-studied. Take during the antibiotic course.
- Acute infectious diarrhea. Shortened duration by ~1 day (the duration analysis pooled a 35-trial subset of the Cochrane review; Allen 2010).
- C. difficile prevention. Reduced risk 4.0%→1.5% overall — but the benefit was concentrated in people at higher baseline C. diff risk; trials in lower-risk patients found no significant effect (Goldenberg 2017).
- IBS — mixed, strain-dependent. Some strains help symptoms, but the evidence isn't conclusive and the ACG guideline recommends against probiotics as a class for IBS (Ford 2018, ACG 2021).
- General "gut health" — weakest. Every solid result above is tied to a specific condition and strain; blanket wellness claims aren't what the good trials tested.
How to take it, and the label catch
- With a meal (or ~30 min before). A meal buffers stomach acid and improved bacterial survival in a simulated-digestion lab model (not a human trial, and tested on one product) — a plausible, low-risk practice (Tompkins 2011).
- During antibiotics, not after. For AAD prevention, take it through the course, a couple of hours apart from the antibiotic dose.
- Give it ~4 weeks to judge a specific strain.
- ⚠️ Watch the CFU wording. Many labels guarantee CFU only "at time of manufacture," and live counts decay over shelf life — testing found a product with zero viable organisms despite a huge labeled count (Taha 2025). Prefer products that guarantee CFU through expiry; refrigerated and spore-forming strains hold up better.
Strain-matched picks
Lactobacillus rhamnosus GG, 10B CFU — the exact strain from the antibiotic-diarrhea trials. Take during the antibiotic course.
Bifidobacterium infantis 35624, the strain studied for IBS symptoms (evidence overall is mixed — see below).
Bacillus coagulans BC30, 2B CFU — a spore-former that survives room temperature well. Lowest cost per day.
Compare all picks on the best probiotics ranking.
Frequently asked questions
How many billion CFU should I take?
Usually 1–10 billion CFU/day (some multi-strain products use 50–100+ billion). But CFU is the wrong shopping metric — the effect is strain-specific, so match the strain to your goal first; the CFU is then whatever that strain's trials used.
Does strain matter more than CFU?
Yes, by a lot. Effects proven for one strain don't transfer to another. The strain code (e.g., L. rhamnosus GG, B. infantis 35624) is what maps to a trial. A big CFU number with only genus names and no condition is marketing, not clinical.
When should I take them, with food?
With a meal or ~30 min before (buffers acid; improved survival in a lab digestion model). For antibiotic-diarrhea prevention, take during the course, a couple hours apart from the antibiotic. Give a product ~4 weeks.
Do probiotics actually work?
For specific strain+condition pairs, yes — strongest for antibiotic-associated diarrhea, acute infectious diarrhea, and C. diff prevention. IBS is mixed (a guideline recommends against class-wide use). General "gut health" claims are the weakest evidence.
Are they safe / do label CFUs last?
Safe for healthy people; NOT for the immunocompromised/critically ill/central-line/short-gut (documented serious infection and mortality risk) — see a doctor. On labels, "CFU at manufacture" decays over shelf life; prefer guaranteed-through-expiry, refrigerated, or spore-forming products.
Related
Sources
- Szajewska H, Kołodziej M. "S. boulardii in the prevention of antibiotic-associated diarrhoea: meta-analysis." Aliment Pharmacol Ther. 2015. PMID: 26216624
- Szajewska H, Kołodziej M. "L. rhamnosus GG in the prevention of antibiotic-associated diarrhoea." Aliment Pharmacol Ther. 2015. PMID: 26365389
- Goldenberg JZ, et al. "Probiotics for prevention of pediatric antibiotic-associated diarrhea." Cochrane. 2015. PMID: 26695080
- Allen SJ, et al. "Probiotics for treating acute infectious diarrhoea." Cochrane. 2010. PMID: 21069673
- Goldenberg JZ, et al. "Probiotics for prevention of C. difficile-associated diarrhea." Cochrane. 2017. PMID: 29257353
- Ford AC, et al. "Prebiotics, probiotics, synbiotics and antibiotics in IBS: meta-analysis." Aliment Pharmacol Ther. 2018. PMID: 30294792
- Lacy BE, et al. "ACG Clinical Guideline: Management of IBS." Am J Gastroenterol. 2021. PMID: 33315591
- Tompkins TA, et al. "Impact of meals on a probiotic during GI transit." Benef Microbes. 2011. PMID: 22146689
- Taha MW, et al. "Bacterial Viability and Label Accuracy in Commercial Probiotics." Microorganisms. 2025. PMID: 40871437
- Besselink MG, et al. "Probiotic prophylaxis in predicted severe acute pancreatitis (PROPATRIA)." Lancet. 2008. PMID: 18279948