We use essential cookies (authentication, your saved goals/stack) by default. With your permission we’ll also enable privacy-respecting analytics (Vercel Web Analytics, anonymous load-time metrics) and error-replay diagnostics (Sentry — DOM snapshots only when an error fires) so we can fix bugs faster. Learn more about cookies
Probiotics work strain by strain, not as a category — so these are the individual, named strains and defined formulations that have their own human trials, rather than a generic "50 billion CFU" blend. L. rhamnosus GG (Culturelle) for antibiotic-associated diarrhoea, B. infantis 35624 (Align/Alflorex) and L. plantarum 299v for IBS symptoms, B. lactis (BB-12/HN019) for gut transit and regularity, L. reuteri (BioGaia) for colic in breastfed infants, and the high-dose VSL#3/Visbiome blend for pouchitis and ulcerative colitis. Each is ranked by the real strength of its evidence, with the indication it was actually tested for — several are far better known for uses where the trials came out null.
Strongest evidence: Saccharomyces Boulardii, VSL#3 / Visbiome, and B. lactis (BB-12 / HN019).
These are specific probiotic strains and defined formulations, not a generic probiotic category — benefit is tied to the exact strain and the dose used in its trials, not to the CFU count on the bottle. Listing reflects evidence strength for the indication each strain was actually tested for; a strain proven for one problem will not necessarily help another. People who are severely immunocompromised or critically ill should check with a clinician before taking live cultures.
Antibiotic-resistant probiotic yeast ideal for preventing traveler's diarrhea and protecting gut flora during antibiotic therapy.
An 8-strain blend dosed in the hundreds of billions to trillions of CFU per day — roughly 45 to 7,000 times a typical 1-10 billion CFU consumer probiotic — with the largest effect size in the probiotic literature for one narrow use: maintaining remission in chronic pouchitis after ulcerative colitis surgery. It is null for IBS and Crohn's — and since 2016 the brand name no longer reliably identifies the formulation that was studied.
The most-replicated probiotic effect on gut transit — it measurably speeds things up and adds roughly one bowel movement a week in people who are already low. But it is not a constipation cure: the two largest, best-designed trials both missed their primary endpoints, and the honest effect is hours off transit time, not a fix.
NOT a capsule you swallow. L. crispatus CTV-05 (LACTIN-V) is a live biotherapeutic inserted into the vagina with an applicator, and every efficacy trial has used that route — there is no oral evidence at all. Used after a course of vaginal metronidazole, it lowered bacterial-vaginosis recurrence from 45% to 30% in the one large trial, but only in women whose infection actually cleared with the antibiotic first.
The best-evidenced probiotic for infant colic — but specifically in breastfed babies, where an individual-participant meta-analysis found a number-needed-to-treat of 2.6. In formula-fed infants the benefit disappears, and the largest single trial actually favoured placebo. Adult claims mostly belong to different L. reuteri strains entirely.
One of the two most-studied probiotic strains, and a case study in why strain-level evidence matters. It has one genuinely strong, guideline-endorsed use — preventing antibiotic-associated diarrhea in children — while its two most-marketed uses, treating stomach bugs and preventing eczema, were both overturned by the largest and best-run trials.
A spore-forming probiotic that survives stomach acid and needs no refrigeration. Its real evidence is in irritable bowel syndrome, where a network meta-analysis of 81 trials ranked the MTCC 5856 strain at the top for abdominal pain and stool form in IBS-D. The immune, protein-absorption and muscle-recovery claims on the label are much weaker than they look, and a meta-analysis was negative for constipation.
The probiotic strain behind Align, and the one most often called 'the evidence-based one' for IBS. The reality is thinner: four single-strain trials, two of them null, sharing overlapping authorship — and the large positive dose-ranging trial enrolled women only. The dose-response is genuinely inconsistent rather than a clean 'less is more'.
The Yakult strain — the most-trialled single probiotic strain of the three psychobiotic-adjacent contenders, and a clear illustration of why trial size matters. The two largest trials (n=737 and n=359) were null on their primary endpoints. The positives cluster in smaller, Yakult-affiliated studies that compared fermented milk against plain milk.
The one probiotic strain with a genuinely distinctive trick: taken with a meal it increases non-heme iron absorption by roughly 50%, replicated across independent dual-isotope studies. Its IBS evidence is far weaker — rated low certainty, with one clearly negative trial — and it does not fix iron deficiency on its own.
The probiotic sold for belly fat. Two Japanese trials and one Korean trial did report less visceral fat, but the Japanese results are within-group changes from baseline rather than proper comparisons against placebo, the effect faded within four weeks of stopping, another human trial found nothing, and the total weight change is around one kilogram.
The 'psychobiotic' strain marketed for stress and sleep. Not the same organism as B. infantis 35624 (Align), despite both being Bifidobacterium longum. After roughly a decade of coordinated research from a single lab, the two largest trials both missed their primary endpoints, and every objective measurement taken — actigraphy, behaviour — was null.
The 'psychobiotic' combination behind Probio'Stick and Cerebiome — always a fixed two-strain pair, never L. helveticus alone. One 110-person trial in diagnosed depression is genuinely positive at p=0.042; an independent 135-person trial in healthy adults was null on every pre-specified outcome; and the widely cited founding trials were either GI-only or reported without sample sizes.