Objective: Lactobacillus GG (L-GG), an acid- and bile-resistant strain that colonizes the intestinal mucosa, has been used to manage diarrhea in children. Our objective was to evaluate the prophylactic use of L-GG to prevent diarrhea in children at high risk from a developing country in a randomized, placebo-controlled trial. Study design: Two hundred four undernourished children 6 to 24 months old from an indigent peri-urban Peruvian town received either L-GG or placebo in flavored gelatin once daily, 6 days a week, for 15 months. Episodes of diarrhea were documented by daily home visits, and diagnostic studies were done in a subset of cases. Recovery of L-GG in stool from subjects and from family contacts was examined. Results: Subjects in the L-GG group had significantly fewer episodes of diarrhea (5.21 episodes diarrhea/child/year [“ecy”] L-GG group, 6.02 ecy placebo group; P = .028). The decreased incidence of diarrhea in the L-GG group was greatest in the 18- to 29-month age group ( P = .004) and was largely limited to nonbreastfed children (Breastfed: 6.59 ecy L-GG, 6.32 ecy placebo, P = .7; Nonbreastfed: 4.69 ecy L-GG, 5.86 ecy placebo, P = .005). The duration of diarrhea episodes and the causes of diarrhea were similar in both groups, except adenovirus was more common in the placebo group. Conclusion: L-GG supplementation may be useful as a prophylactic measure to control diarrhea in undernourished children at increased risk, especially nonbreastfed children in the toddler age group. (J Pediatr 1999;134:15-20.)