Probiotics for Kids: The 3 Strains With Trials (and the Rest)
Not medical advice. This summarizes the Cochrane reviews and randomized trials in children, the labels manufacturers file with the NIH, and current Amazon prices. A probiotic for a premature, immunocompromised or seriously ill child is a specialist’s decision.
Quick answer
Three strains have pediatric trials, and most of the children’s shelf does not carry them. For a child on antibiotics, probiotics cut antibiotic-associated diarrhea from 19% to 8% across 33 trials, with L. rhamnosus GG and S. boulardii the best-studied and 5 billion CFU a day the dose that worked. For a breastfed baby with colic, L. reuteri DSM 17938 cut crying by 25.4 minutes a day. For a stomach bug, the best trials found nothing: a 5-day course of GG in 943 preschoolers did no better than placebo.
Buy the strain for the problem, not a “kids’ daily” gummy. On the 149 children’s probiotic labels filed with the NIH, 8.1% name L. rhamnosus GG, 0% name L. reuteri DSM 17938, and 49.7% give any strain identifier at all. Culturelle Kids Purely Probiotics Chewables is GG at 5 billion CFU for $0.68 a dose; BioGaia Protectis Baby Drops is DSM 17938 for $0.70 a dose. See the census →
Choosing for an adult? See the strain matcher and the ranked probiotics →
On antibiotics: the one use with a strong trial base
Antibiotic-associated diarrhea is the clearest case. The Cochrane review of probiotics for preventing it in children pooled 33 trials with 6,352 participants: diarrhea occurred in 8% of children given a probiotic against 19% on placebo or no treatment, a number needed to treat of 9, on moderate-certainty evidence. Doses of 5 billion CFU a day or more were more effective than lower doses, and the best-studied organisms were Lactobacillus rhamnosus GG and Saccharomyces boulardii; no serious adverse events were attributed to probiotics in the trial children (Guo 2019, PMID 31039287). The practical rule: start the probiotic with the first antibiotic dose, keep it up for a few days after the course ends, and take it a couple of hours apart from the antibiotic if it is a bacterial strain (S. boulardii is a yeast and does not care).
Colic: one strain, breastfed babies, a large effect
The colic evidence belongs to one organism. An individual-participant meta-analysis of 4 double-blind trials in 345 infants with colic found L. reuteri DSM 17938 reduced crying and fussing by 25.4 minutes a day at day 21 against placebo, with treatment success about 1.7 times as likely. The effect was concentrated in breastfed infants, where the number needed to treat was 2.6; in formula-fed infants it was not significant and the data were too thin to conclude (Sung 2018, PMID 29279326). The trial dose was five drops a day, 100 million CFU, which is a fraction of what the shelf advertises and a reminder that the strain carries the evidence, not the count.
Stomach bugs: the biggest trials found nothing
Acute gastroenteritis is where parents reach for a probiotic and where the evidence turned against it. The 2020 Cochrane update pooled 82 trials in 12,127 people, 11,526 of them children; in the trials at low risk of bias there was no difference between probiotic and control in the risk of diarrhea lasting 48 hours or more, and the authors could not say whether probiotics shorten diarrhea at all (Collinson 2020, PMID 33295643). The trial that moved the field was American: 943 children aged 3 months to 4 years presenting to emergency departments with gastroenteritis were randomized to five days of L. rhamnosus GG or placebo, and 11.8% versus 12.6% went on to moderate-to-severe illness (relative risk 0.96), with diarrhea lasting a median 49.7 against 50.9 hours (Schnadower 2018, PMID 30462938). Oral rehydration solution is the treatment for a child’s stomach bug; the electrolyte guide covers the glucose-paired formulas built for it.
We read 149 children’s probiotic labels: half name no strain at all
Original research259 labels retrieved from the NIH Dietary Supplement Label Database by paging its search for children’s, kids’, toddler and infant probiotic terms and keeping labels with a probiotic organism whose name says kid, child, toddler, infant, baby or junior; 149 sold as probiotics (110 multivitamin-with-probiotic and other lines excluded) from 83 brands, 3,741 probiotic labels of any kind on the market. Every figure was recomputed in two halves; 11 of 19 held and only those are printed.
49.7% of children’s labels give any strain identifier; the median label names 2 species. The trial strains are rare: 8.1% name L. rhamnosus GG, 6% S. boulardii, 2% BB-12, and 0% name L. reuteri DSM 17938 in the children’s lines we could enumerate (the colic product is sold as an infant drop and filed under its own brand). 6.7% of lines are drops or liquids and 10.7% are infant lines; 42.3% add a prebiotic and 25.5% claim sugar-free. CFU is invisible to the database: only 3.4% of labels (2) print a colony count as a structured quantity, so no CFU distribution can be computed from labels, and the shelf’s counts live on the front of the box.
| Item | Value (%) |
|---|---|
| Any strain identifier given | 49.7% |
| Prebiotic (FOS/inulin/HMO) added | 42.3% |
| L. rhamnosus GG named | 8.1% |
| S. boulardii named | 6% |
| BB-12 named | 2% |
| L. reuteri DSM 17938 named | 0% |
How we read the labels, and what did not hold
Method: organisms and strain identifiers are read from Supplement Facts rows, declared forms and the ingredient statement (a strain ID is an alphanumeric code such as GG, ATCC 53103, DSM 17938, BB-12, CNCM I-745); format from the filed physical state and product name; prebiotics from FOS, inulin, GOS or HMO in any row. Split-half by label-id parity: 11 of 19 checks agreed. Not printed (8): median CFU (billions, structured only); CFU at least 5 billion share (structured); CFU at least 10 billion share (structured); B. infantis share; any pediatric-trial strain share; single-species share; gummy share; sugars declared share. The census is published as open data at /probiotics/for-kids.json (CC BY 4.0). What it cannot tell you: CFU on almost any label, whether the organisms are alive at the end of shelf life, or anything about a brand that has not filed with the NIH.
Which probiotic to buy for a child
Product links go to Amazon and we may earn a commission if you buy. It never changes which product we pick or what we say about it. How we choose.
What not to buy: a daily “kids’ probiotic gummy” for a healthy child. No trial shows a benefit in a child with nothing wrong, the gummy format is the one live organisms tolerate worst, and the strains that survive it are rarely the ones with pediatric trials.
Safety, briefly
In healthy children the trial strains were well tolerated, with no serious adverse events attributed to probiotics across the pediatric antibiotic-diarrhea trials (Guo 2019). The same review is explicit about who is different: observational reports of probiotic bacteria and yeast reaching the bloodstream have come from severely debilitated or immunocompromised children, including those with central venous catheters or conditions that let organisms cross the gut wall. Premature infants, children on chemotherapy or immunosuppressants, and children with central lines get probiotics only on a specialist’s instruction. S. boulardii is a yeast and should not be handled near a central line at all. For everyone else, the risks are the mundane ones: sugar in gummies, and money spent on a species without a strain.
Frequently asked questions
What is the best probiotic for kids?
The one with the trial for the problem you have. For a child on antibiotics, Lactobacillus rhamnosus GG (Culturelle Kids Purely Probiotics Chewables, 5 billion CFU a chewable) or Saccharomyces boulardii (Florastor Kids Probiotic Powder): across 33 trials, probiotics cut antibiotic-associated diarrhea from 19% to 8% of children, and doses of 5 billion CFU a day or more worked best. For a breastfed baby with colic, L. reuteri DSM 17938 (BioGaia Protectis Baby Drops), which cut crying by 25.4 minutes a day in an individual-patient meta-analysis. For a healthy child with no problem, there is no trial showing a daily probiotic does anything.
Do probiotics help a child's stomach bug?
The best evidence says no. The Cochrane review of 82 trials in 12,127 people, 11,526 of them children, found no difference between probiotic and placebo in the risk of diarrhea lasting 48 hours or more in the low-bias trials, and a large US emergency-department trial of L. rhamnosus GG in 943 preschool children found 11.8% with moderate-to-severe illness on the probiotic against 12.6% on placebo, with diarrhea lasting 49.7 against 50.9 hours. Oral rehydration is the treatment; a probiotic is not.
How many CFU should a kids' probiotic have?
For antibiotic-associated diarrhea, the Cochrane subgroup found 5 billion CFU a day or more more effective than lower doses. For colic the trial dose was 100 million CFU of one strain, which shows that the strain, not the count, is what the evidence attaches to. The NIH label database cannot see CFU except when a label prints it as a structured quantity, which 3.4% of children's labels do (2 labels), so a CFU comparison of the shelf is not possible from labels; the number on the front of the box is the only source, and it is the count at manufacture, not at the end of shelf life.
Are probiotic gummies for kids any good?
A gummy is the format least suited to a live organism: it is cooked, moist and sugared, and the strains that survive it are usually spore-formers (Bacillus) rather than the Lactobacillus and Saccharomyces strains in the pediatric trials. On the children's labels we read, the gummy share did not hold across halves and is not printed; what did hold is that 25.5% of children's probiotics claim sugar-free. A chewable tablet, a powder stick or drops delivers the trial strains without the gel.
Is a probiotic safe for a baby?
The trial strains have good safety records in healthy term infants: L. reuteri DSM 17938 in the colic trials from the first weeks of life, L. rhamnosus GG in thousands of children. The exceptions are real and specific: premature or immunocompromised infants, children with central venous catheters, and children with short-gut or severe illness, where Cochrane notes case reports of probiotic bacteria and yeast reaching the bloodstream. Those children get probiotics only under a specialist's direction.
What does "strain" mean, and why does it matter?
A species name (Lactobacillus rhamnosus) is a family; a strain (L. rhamnosus GG, ATCC 53103) is the specific organism a trial used, and benefits do not transfer between strains. On the 149 children's labels we read, 49.7% give any strain identifier; the rest sell a species and a count. Match the strain on the label to the strain in the trial: GG for antibiotic diarrhea, DSM 17938 for breastfed colic, S. boulardii CNCM I-745 as the yeast alternative.
Related guides
- Probiotics: matching strains to problems
- Probiotic strain matcher
- Best probiotic for antibiotic diarrhea
- Probiotics for women
- Best probiotics, ranked
- Electrolyte powders (and the oral-rehydration formula for a stomach bug)
Sources
- Guo Q, et al. “Probiotics for the prevention of pediatric antibiotic-associated diarrhea.” Cochrane Database Syst Rev. 2019. PMID: 31039287
- Sung V, et al. “Lactobacillus reuteri to Treat Infant Colic: A Meta-analysis.” Pediatrics. 2018. PMID: 29279326
- Collinson S, et al. “Probiotics for treating acute infectious diarrhoea.” Cochrane Database Syst Rev. 2020. PMID: 33295643
- Schnadower D, et al. “Lactobacillus rhamnosus GG versus Placebo for Acute Gastroenteritis in Children.” N Engl J Med. 2018. PMID: 30462938
- NIH Office of Dietary Supplements. Dietary Supplement Label Database. dsld.od.nih.gov (labels retrieved 2026-09-18).