A randomized clinical trial was completed to study the severity, duration, and nutritional outcome of acute diarrhea in 85 Peruvian children between 5 and 24 months of age who received a soy-protein-isolate, lactose-free formula (group SF) or one of two mixtures of home-available foods, all in amounts up to 110 kcal/kg of body weight/day. The mixed diets contained either wheat flour, pea flour, carrot flour, sucrose, and vegetable oil (group WP) or potato flour, dried whole milk, carrot flour, sucrose, and oil (group PM). The characteristics of the children in each group were generally similar initially. There were no differences in treatment failure rate by diet group. Fecal outputs were similar in all groups during the first day of treatment, averaging 60 to 65 g/kg/day. How-ever, stool outputs were greater for the PM group than for other groups on days 3 and 4, and were less for the SF group than the other groups on days 5 and 6. The esti-mated median durations of diarrhea in the WP group (52 h) and PM group (53 h) were significantly less than in the SF group (154 h, p = 0.005). Energy intakes, energy ab-sorption, and nitrogen retention (% of intake) were gen-erally similar in all dietary groups, although there were minor differences in the absorption of specific macronu-trients. Children in all groups gained weight during hos-pitalization, and there were no significant differences by diet group in the change in anthropometric status during treatment. We conclude that these locally available, low-cost staple food mixtures offer a safe and nutritionally adequate alternative to a commercially produced lactose-free formula for the dietary management of young children with acute diarrhea in this setting. © 1991 Raven Press Ltd, New York.