Planning guide
How to Choose a Probiotic
Short answer: There's no single best probiotic. What a probiotic does depends on the exact strain, the dose and what you're taking it for. [S02] [S03] The strongest evidence is for a few specific uses, such as preventing antibiotic-associated diarrhea with Lacticaseibacillus rhamnosus GG or Saccharomyces boulardii. [S03] To choose well, start with your reason, then pick a product that names the full strain. Look for one that lists live CFUs through the expiration date and matches a dose that's been studied. [S02] [S03] People who are seriously ill or have weakened immune systems should talk to a clinician before taking one, because their risk of harmful effects is higher. [S01]
Probiotic shelves are crowded with "billions of CFUs," long ingredient lists and gut-health promises. Most of that information doesn't tell you whether the product will help. This guide explains how to read a probiotic label, what the evidence supports and who should be careful. If you take prescription medicines or other supplements, read our herb-drug interactions guide too.
Start with why you want one
NIH's Office of Dietary Supplements says there are no official recommendations for probiotic use in healthy people. [S02] It suggests asking a health care provider which product, dose and duration to use. [S02] That matters because probiotics aren't interchangeable. Some mechanisms are shared by many strains, but others are unique to one species or strain. [S03]
Here's what the research summarized by NIH shows for common reasons people buy them:
Reason | What the evidence shows | Strength |
|---|---|---|
Preventing diarrhea from antibiotics | LGG or S. boulardii started within 2 days of the first dose lowered risk in children and adults 18 to 64 [S03] | Moderate for children [S04]; no benefit seen in adults 65+ [S03] |
Irritable bowel syndrome (IBS) | Modest benefit for some symptoms in meta-analyses; effects vary by strain and symptom [S03] | Low-quality studies [S03] |
Ulcerative colitis | Might slightly reduce symptoms alongside medication [S02] | Limited; the American Gastroenterological Association (AGA) advises use only in clinical trials [S03] |
Crohn's disease | No evidence of benefit [S03] | Not supported |
Acute infectious diarrhea in children | Mixed; the AGA recommends against routine use in the US and Canada [S03] | Conflicting guidelines [S03] |
High cholesterol | Some studies show a slight drop in total and LDL cholesterol; others show no benefit [S02] | Mixed |
General "gut health" in healthy people | No official recommendations [S02] | Not established |
If your reason isn't on the list, or the evidence is weak, a probiotic may not be the right tool. For ongoing digestive symptoms, a diagnosis comes first.
Read the strain, not just the brand
Probiotics are named by genus, species and strain. In Lacticaseibacillus rhamnosus GG, Lacticaseibacillus is the genus, rhamnosus is the species and GG is the strain. [S02] Some strains also have a subspecies name and an alphanumeric code. [S03]
The strain is the part that connects a product to research. A study on one strain of L. rhamnosus doesn't prove that a different strain of the same species works the same way. [S02] [S03] A label that only says "Lactobacillus blend" or "proprietary probiotic complex" makes it hard to know which evidence, if any, applies.
You may also notice that familiar names have changed. Several species that used to be in the Lactobacillus genus were moved to new genera when the genus was restructured in 2020. [S03] Older studies and labels may still use the old names.
What to look for: the genus, species and strain code for every organism in the product. [S02]
Understand CFU counts
Probiotic amounts are measured in colony-forming units (CFUs), the number of live, active microorganisms. [S02] Labels often show them as 1 x 10⁹ (1 billion) or 1 x 10¹⁰ (10 billion). [S02] Many supplements contain 1 to 10 billion CFUs per dose, and some contain 50 billion or more. [S03]
A few points help make sense of the numbers:
- Bigger isn't automatically better. Higher CFU counts alone don't mean greater health benefits. The strain and dose together matter. [S02] [S03]
- Dose still counts for some uses. In a Cochrane review of children taking antibiotics, doses of 5 billion CFUs or more per day worked better than lower doses. [S04]
- Check the date the count applies to. Probiotics are alive and can die during storage. Some labels list CFUs at manufacture, others through the expiration date. Experts suggest choosing products that guarantee the count through the expiration or use-by date. [S02] [S03]
- Weight isn't the same as live count. Federal rules only require the total weight of microorganisms on the Supplement Facts label. That weight can include dead cells. Listing CFUs is voluntary. [S03]
What the antibiotic evidence actually shows
Antibiotics can disrupt the gut's microbes, and diarrhea is a common side effect. [S03] This is the use with the most consistent evidence.
A 2019 Cochrane review pooled 33 randomized trials in 6,352 children taking antibiotics. [S04] Diarrhea occurred in 8% of children given probiotics compared with 19% of those who weren't. That means about 9 children would need to take probiotics to prevent one case. [S04] The reviewers rated the evidence as moderate certainty. Twenty of the 33 trials were at high risk of bias, and the probiotics tested varied widely. [S04]
In adults, NIH's summary of meta-analyses found that LGG and S. boulardii both reduced the risk. [S03] Starting within 2 days of the first antibiotic dose worked better than starting later. [S03] The benefit was seen in adults 18 to 64 but not in those 65 and older. There's no evidence that products with more than one strain work better. [S03]
If you want to try one with an antibiotic, ask your prescriber or pharmacist which strain and dose to use and how to time it.
Check the rest of the label
Beyond strain and CFUs, a few label details are worth a look:
- Storage instructions. Some probiotics need refrigeration and others are fine at room temperature. Store them as directed. [S02]
- Expiration or use-by date. Because CFUs decline over time, a product near its date may contain fewer live organisms. [S02]
- Serving size. Make sure the CFU count applies to one serving, not several capsules.
- Health claims. Supplement labels can describe effects on the body's structure or function without FDA approval. They can't claim to lower your risk of a disease without FDA consent. [S01]
Supplements vs fermented foods
Yogurt, kefir, sauerkraut, kimchi and other fermented foods are made with live microbes. But that doesn't mean they contain probiotics with proven health benefits. [S02] If a food has added probiotics, the label should list the genus, species and strain. [S02]
Fermented foods can be part of a healthy diet. Just don't assume they'll match a studied probiotic dose or strain.
How probiotics are regulated
Depending on how it's marketed, the FDA may regulate a probiotic as a dietary supplement, a food ingredient or a drug. [S01] Many are sold as dietary supplements, which don't need FDA approval before they go on sale. [S01] A probiotic marketed to treat a disease would need to go through clinical trials and FDA approval, like a drug. [S01]
Quality varies. NCCIH notes that some probiotic products have been found to contain microorganisms not listed on the label. In some cases those contaminants can pose serious health risks. [S01]
Who should be careful
Most probiotics have been used safely in foods and supplements for decades, and in healthy people the most common side effect is gas. [S02] But detailed safety research is limited, so the frequency and severity of side effects aren't well known. [S01]
The risks are higher for some groups:
- Preterm infants. The FDA has warned that preterm infants given probiotics in hospitals are at risk of invasive, potentially fatal infections. It cited one infant death in 2023 and more than two dozen other reported adverse events since 2018. [S05] The FDA hasn't approved any probiotic as a drug or biological product for infants. [S05]
- People who are seriously ill or have weakened immune systems. Harmful effects are more likely in these groups, and the potential risks should be carefully weighed against benefits. [S01] The World Gastroenterology Organisation advises limiting use in these groups to strains and uses with proven benefit. [S03]
- People in intensive care or with central lines. Serious infections linked to probiotics are rare, but NIH notes that in many reported cases of bloodstream infections linked to Saccharomyces yeast probiotics, the patients were in intensive care, on tube or IV feeding, had a central venous catheter or were on broad-spectrum antibiotics. [S03] In the Cochrane review of otherwise healthier children, none of the 24 trials that tracked side effects found serious problems caused by probiotics. [S04]
NCCIH advises talking with a health care provider before starting a probiotic, especially if you have health problems. Anyone with a serious underlying condition should be monitored closely. [S01] If you're not sure who to ask, our guide to what a naturopathic doctor does explains how that kind of visit works.
A quick checklist
- Name your goal and check whether a probiotic has evidence for it.
- Choose a product that lists the full strain for each organism. [S02]
- Match the strain and dose to what was studied for your goal. [S03]
- Pick a CFU count guaranteed through the expiration date. [S02] [S03]
- Follow the storage instructions. [S02]
- Check with your clinician if you're pregnant, immunocompromised, seriously ill or buying for an infant. [S01] [S05]
- Give it a fair, defined trial, and stop if it isn't helping.
Questions
What is the best probiotic to take?
There isn't one best probiotic. Benefits depend on the exact strain, the dose and the reason you're taking it. For preventing diarrhea from antibiotics, the best-studied strains are Lacticaseibacillus rhamnosus GG and Saccharomyces boulardii. For most other uses the evidence is mixed or weak, so match the product to a use that has actually been studied.
How many CFUs should a probiotic have?
There's no single right number. Many supplements contain 1 to 10 billion CFUs per dose, and a higher count doesn't automatically mean a better product. Look for the CFU count guaranteed through the expiration date, and for a dose close to what was used in studies of that strain.
Should I take a probiotic with antibiotics?
It may help some people avoid antibiotic-associated diarrhea. NIH's Office of Dietary Supplements says starting LGG or Saccharomyces boulardii within 2 days of the first antibiotic dose reduced risk in children and adults aged 18 to 64, but not in studies of adults 65 and older. Ask your prescriber or pharmacist first.
Are multi-strain probiotics better?
Not necessarily. For antibiotic-associated diarrhea, NIH's Office of Dietary Supplements found no evidence that more than one strain works better. Some IBS studies favored multi-strain products, but the research was low quality. More strains on the label doesn't mean more benefit.
Who should not take probiotics?
Preterm infants, people who are seriously ill and people with weakened immune systems face a higher risk of serious infection from probiotics. The FDA has warned about deaths and infections in hospitalized preterm infants. Anyone in these groups should only use a probiotic under medical supervision.
Related guides
Apple Cider Vinegar: What It Can and Can't Do
Vinegar with a meal modestly lowers the blood sugar spike afterward, and small trials suggest small effects on weight and fasting glucose. The evidence is short-term and mixed, and the most-cited weight-loss trial was retracted in 2025.
Read the guide: Apple Cider Vinegar: What It Can and Can't DoAshwagandha: Benefits, Side Effects and Evidence
Small trials suggest some ashwagandha extracts may ease stress and modestly improve sleep, but the studies are short, mostly from India and vary in product. It is usually well tolerated for up to three months, with rare liver injury and possible thyroid effects.
Read the guide: Ashwagandha: Benefits, Side Effects and Evidence
Spotted an error or an outdated source? Tell us which page and what changed. Please do not include sensitive information.
Report a correction