A randomized ambulatory trial was performed to compare early (6-h) vs. late (24-h) refeeding in acute infantile diarrhea. Ninety infants with mild dehydration were enrolled in the study. Following an initial oral rehydration period (WHO formula), refeeding was introduced using a diet based on either breast milk or cow's milk. Early (n = 53) and late (n = 37) refeeding groups were similar in ethnic background, socioeconomic level, relevant past history, nutritional and clinical state, and stool pathogens. Infants were assessed upon their initial visit, at 24 and 48 h, and at 7 and 14 days thereafter for evaluation of weight, hydration state, stool frequency and need of hospitalization. No significant differences in the above parameters were observed between the two groups. Different patterns of refeeding (breast milk vs. cow's milk) in both early and late refeeding groups showed no significant differences in the features studied. Since the short-term clinical outcome following early refeeding in acute infantile diarrhea is not different from late refeeding, we suggest that early refeeding should be preferred, particularly in developing populations, in order to minimize the adverse nutritional effects of prolonged fasting during recurrent bouts of gastroenteritis.