Saccharomyces Boulardii

The verdict

Is Saccharomyces Boulardii worth taking?

This is the compound in the guide where a single grade is least useful. The evidence is strong for some indications, very weak for others, and null for the one it is most often bought for. The grade of Strong reflects the best of it, and the rest of this page is the breakdown that the grade cannot carry on its own.

Where it earns the grade: preventing antibiotic-associated diarrhea. In children, a Cochrane review of 33 trials and 6,352 participants put the incidence at 8 percent with probiotics against 19 percent without, with moderate-certainty evidence and a number needed to treat of nine. At doses of five billion CFU a day or more the number needed to treat falls to six. S. boulardii is one of the two organisms the reviewers name as most appropriate. In adults the picture is similar. This part is real and replicated.

Where it does not: preventing Clostridioides difficile. The strain-specific meta-analysis — eight studies, 9,712 patients for primary prevention — found an odds ratio of 0.71 with a confidence interval of 0.46 to 1.10 and a p-value of 0.124. No benefit. For preventing recurrence, two studies and 292 patients: 36.7 percent against 46.2 percent, also not significant. The authors' conclusion is that S. boulardii appears to have no benefit for preventing either primary or recurrent C. difficile infection in people taking systemic antibiotics.

The reason almost every page says otherwise is that they cite reviews that pool all probiotics together. Those pooled analyses do show a benefit — the 2025 Cochrane review covers 47 studies and 15,260 participants. But pooling many organisms and reading the result back onto one of them is not how this works, and when someone has done the strain-specific analysis, it comes out null. The guidelines disagree with each other too: the American Gastroenterological Association conditionally supports it on low-certainty evidence, while the American College of Gastroenterology recommends against it for recurrence prevention, and does so strongly.

The rest, briefly. Acute gastroenteritis in children: about a day shorter, from a body of trials the reviewers themselves rate very low quality. H. pylori treatment: a small eradication gain, around six to twelve percentage points, but a large and consistent reduction in the side effects of the antibiotics, with diarrhea roughly a third as common. Traveler's diarrhea: a modest benefit from Austrian tourist trials run in the 1980s and 90s and never replicated since. Irritable bowel syndrome and Crohn's: tested properly and negative. SIBO and candida: no randomized trial exists.

And one thing the marketing gets right: it does not colonize. It passes through, reaching steady state in stool within about three days and clearing within three to five days of stopping. There is no building it up and no tapering off. You take it through the period you want it, or not at all.

What it does

What Saccharomyces Boulardii does in the body

Antibiotic-associated diarrhea prevention, C. difficile infection and recurrence reduction, traveler’s diarrhea prevention, and gut microbiome resilience during and after antibiotic treatment. Saccharomyces boulardii is a tropical yeast probiotic that occupies a unique position in the probiotic landscape: being a fungus rather than a bacterium, it is completely unaffected by antibacterial antibiotics and can colonize the gut and confer protective effects during antibiotic courses when all bacterial probiotics are simultaneously being eliminated. It also demonstrates activity against pathogenic bacteria through competitive exclusion, toxin binding, and immune modulation, all without establishing permanent colonization since it clears within 3 to 5 days of stopping supplementation.

The evidence

What the research actually shows

Laid out indication by indication, because the spread is the whole story:

Indication	Best evidence	Result
Antibiotic-associated diarrhea, children	Cochrane review, 33 RCTs, 6,352 children (all probiotics)	8% vs 19%, RR 0.45. NNT 9, or 6 at ≥5 billion CFU/day. Moderate certainty
Antibiotic-associated diarrhea, S. boulardii specifically	Meta-analysis, 21 studies	RR 0.47. Absolute risk reduction 8.5 to 18.7%
Antibiotic-associated diarrhea, adults	Meta-analysis, 10 RCTs	RR 0.47. NNT 10
C. difficile, primary prevention, S. boulardii specifically	Meta-analysis, 8 studies, 9,712 patients	OR 0.71 (0.46–1.10), p=0.124. No benefit
C. difficile, recurrence	2 studies, 292 patients	36.7% vs 46.2%, p=0.19. No benefit
Acute gastroenteritis, children	Meta-analysis, 29 RCTs, 3,450 children	Diarrhea about 1.06 days shorter. Very low quality
H. pylori adjunct — eradication	Meta-analysis, 19 trials, 5,036 patients	RR 1.11. Roughly 6 to 12 points absolute. Low certainty
H. pylori adjunct — side effects	Same analyses	Total adverse events RR 0.49; diarrhea RR 0.36. The larger effect
Traveler’s diarrhea	Meta-analysis, 11 RCTs	RR 0.85. Trials are Austrian, 1980s–90s, unreplicated
Irritable bowel syndrome	1 RCT, 67 patients, 4 weeks	Quality of life improved; symptom scores and stool did not
Crohn’s, maintaining remission	1 RCT, 165 patients, 52 weeks	Relapse 47.5% vs 53.2%. Negative
Ulcerative colitis	1 open-label pilot	Not an evidence base
SIBO	No randomized trial	One open, non-randomized pilot in systemic sclerosis
Candida overgrowth, immunity, weight, skin, mood	No human trial found	—

Two reading notes. Where a row says "all probiotics," the trial population took a mixture of organisms and the result cannot be assigned to this one — that distinction is exactly what flips the C. difficile conclusion. And a GRADE rating of very low, as on pediatric gastroenteritis, is the reviewers' own judgment, not a critic's: in that body of work only 38 percent of trials adequately generated their randomization and 17 percent adequately concealed allocation.

Worth checking

It is not a separate species. It is baker’s yeast.

Saccharomyces boulardii is sold as its own organism, with its own name, as though it were something exotic found on tropical fruit. Genetically it is a strain of Saccharomyces cerevisiae — ordinary baker's and brewer's yeast. The correct name is Saccharomyces cerevisiae var. boulardii, and the standalone species name has been formally invalid since 1998.

The evidence is not subtle. A 1998 study found three commercial strains gave patterns identical to S. cerevisiae across all ten restriction enzymes tested, while nine of ten other recognized Saccharomyces species were distinguishable by the same method. A 2003 study found the two identical across the standard identification regions. Whole-genome sequencing of five commercial strains from five different manufacturers in 2017 put average nucleotide identity to S. cerevisiae above 99 percent, and the authors wrote that there is no doubt it belongs to that species. The conventional boundary between two species sits around 95 percent.

What is genuinely different is small and specific. It carries an extra copy of one chromosome. It cannot complete sporulation. It has a reduced copy number of a copper-handling gene and is measurably more copper-sensitive. It lacks intact copies of several transposable elements. And it does survive better at pH 2 to 3, which matters for something that has to get through a stomach.

One widely repeated difference does not hold up. The claim that it grows optimally at human body temperature while ordinary baker's yeast prefers 30 degrees appears in reviews everywhere, but the most recent head-to-head comparison found growth rates at 37 degrees comparable between the two, and most S. cerevisiae strains growing faster across the conditions tested. There is no primary growth-curve data supporting the 37-versus-30 claim that I could find.

The part you can act on: the strain code. Commercial lineages are near-identical to each other genetically but they are not the same deposit, and performance varies with how they are manufactured and grown. Essentially all of the positive trial evidence is on CNCM I-745. A bottle that says only "Saccharomyces boulardii," with no strain designation, is not the organism those trials tested. CNCM I-745 printed on a label is a specific, checkable claim. Its absence is also informative.

The safety point nobody mentions

Fungemia, and why hospitals do not open the capsules in the room

This is a live yeast, and in the wrong patient it can get into the bloodstream. The risk is not spread evenly, and the people it concerns are identifiable.

A 2005 review found 92 published cases of invasive Saccharomyces infection, with S. boulardii accounting for just over half of the fungemias. A Finnish study across five university hospitals over ten years found 46 patients with Saccharomyces fungemia, of whom at least 43 percent were taking S. boulardii; mortality was 22 percent at day seven and 37 percent at day 28. The dominant risk factor across all of these series is an indwelling central venous catheter, followed by critical illness, immunocompromise, and compromised gut integrity. Note that it is the catheter rather than the immune system doing most of the work here: S. boulardii fungemia patients were often immunocompetent.

And it does not only reach the person taking it. In an eight-bed intensive care unit in Rome, four patients developed S. cerevisiae var. boulardii fungemia and none of them had received the probiotic. Gel electrophoresis matched the clinical isolates to the probiotic preparation. The transmission route was a staff member opening and mixing the sachets at a sink about three meters from the nearest bed, without changing gloves between patients. Earlier work had already shown that opening a packet contaminates the air, surfaces and hands, that organisms persist on surrounding surfaces for up to two hours, and that hands stay contaminated despite vigorous washing.

European regulators acted on this; American ones did not have to. In 2017 the EU required S. boulardii labeling to carry a contraindication in patients with a central venous catheter and in critically ill or immunocompromised patients, following ten fatal cases where a causal role could not be excluded — including some without a central line. The same decision requires that sachets or capsules not be opened in patient rooms, and that staff wear gloves and discard them immediately.

In the United States this is a dietary supplement and needs no pre-market approval. No retail label I have seen carries the handling instruction. If you have a family member in a hospital bed with a central line, this is worth knowing, and it is the one thing on this page that could matter in a day rather than over months.

No reliable rate of fungemia per person taking it exists, because none of the case series has a denominator. For a healthy person with an intact gut and no central line, the documented risk is very low. The point is not that this is dangerous. It is that it is a live organism, it is treated as a medicine with a hard contraindication in Europe and as a grocery item here, and nobody tells the person buying it.

Reading the label

Milligrams are not CFU, and the gap is twofold

Trials report doses in two currencies and products print only one of them, which makes dose-matching harder than it looks.

The doses that worked: at least five billion CFU a day for antibiotic-associated diarrhea in children, with reviewers naming five to forty billion as the appropriate range; 250 to 750 mg a day for five or six days in acute gastroenteritis; 500 mg a day for fourteen days alongside H. pylori treatment; two to twenty billion CFU a day in the traveler's diarrhea trials.

Now the conversion problem. The leading branded product states 250 mg as 2.5 billion CFU and 500 mg as 5 billion — ten billion per gram. But a 2021 study in Clinical Infectious Diseases describes its 500 mg twice-daily protocol as twenty billion CFU a day, which is twenty billion per gram. The same milligram figure means a twofold different number of live organisms depending on the source. A reader comparing two "250 mg" products cannot infer potency from the label, and the standard US drug monograph gives no CFU figure at all.

The practical rule: find the CFU count, not the milligrams. If a product does not print one, it has not told you the dose.

On taking it alongside antibiotics — the advice is almost certainly right and is less well-evidenced than it sounds. The reasoning is that antibacterial drugs should have no effect on a eukaryote, and that is sound. But I could not find a primary study demonstrating viability or stool recovery during concurrent antibiotic therapy. It is mechanistically solid and empirically unsourced, which is a fair thing to know about a claim printed on the front of the box.

On timing, the weak evidence points toward starting early. In the retrospective hospital study, starting within 24 hours of the antibiotic performed better than starting later, and the guideline bodies that discount one large negative trial do so partly because it started probiotics three to seven days in. Given that stool levels reach steady state in about three days and clear in three to five, starting with the antibiotic and continuing through the course is the reading that fits the pharmacokinetics.

Food first

Getting Saccharomyces Boulardii from food

You cannot get it from food in any useful amount, and the reason is more interesting than the usual "take the supplement instead."

Since this organism is a strain of Saccharomyces cerevisiae, the species is already all around a kitchen — it is the yeast in bread, beer and wine. But the baking yeast in a loaf is a different strain, and it is dead by the time the bread comes out of the oven. The live yeast in unfiltered beer or in kombucha is not this strain either. None of the acid tolerance or the clinical trial record transfers.

Nutritional yeast and brewer's yeast are both S. cerevisiae and both are deactivated. They are a decent source of B vitamins and protein and they do nothing probiotic.

The honest food-first position for gut health generally — fermented foods with live cultures, a varied range of plant fiber, enough of it to actually matter — is well supported and is a different intervention from this one. S. boulardii is a short-term, targeted tool: you take it through a course of antibiotics or a week of travel, and then you stop, because it washes out in three to five days regardless.

One kitchen note worth having. If you are opening sachets to mix into food or drink, do it away from anyone who is unwell, and wash your hands after. That is not fussiness; it comes from an intensive care unit where four people who never took the product were infected by it.

What the guide adds

The part that is not on this page

Everything above is the assessment. What the guide adds for Saccharomyces Boulardii is the protocol:

• Optimal Form
• Bioavailability
• Purity Red Flags
• Stack With / Avoid

147 entries across nine categories, every one graded against the clinical literature the same way.

Get The Complete Supplement Guide →

About the author

Who graded this

Daniel Edlund is a Le Cordon Bleu-trained private chef in Los Angeles, a certified personal trainer and strength coach, and an internationally licensed massage therapist. Thirty years of strength training and fifteen years cooking in private estate kitchens. He is the author of The Complete Supplement Guide, which grades 147 entries against the clinical literature, and of The EASY Way to Organic Cooking, first published in 2013.

More about Daniel →

This page is information, not medical advice. Supplements interact with prescription medication and with each other, and the right answer depends on your own bloodwork and history. Talk to your doctor before starting anything, particularly if you are pregnant, nursing, managing a health condition, or already taking medication.

Also in Gut Health & Digestion

Related entries

• Psyllium Husk — Strong evidence
• Berberine — Strong evidence
• Probiotics — Moderate evidence
• Prebiotics — Strong evidence
• Postbiotics — Moderate evidence
• Apple Cider Vinegar — Emerging evidence

All 147 entries, graded →