Several recent clinical trials have examined the optimal timing of feeding and the appropriate foods for use in the dietary management of acute childhood diarrhea. Important methodologic issues relevant to these trials, such as the composition and masking of study diets, measurement of outcome variables, and assessment of nutritional status, are discussed. Available data suggest that feeding should be continued during diarrhea, although a small subgroup of patients receiving nonhuman milk exclusively may have an excess rate of complications. These children should be closely supervised or provided with alternative diets. By contrast, children receiving human milk, nonmilk formulas, and mixed diets are generally able to tolerate these foods and to benefit nutritionally from continued feeding.