Controversy continues regarding the ideal composition of glucose/electrolyte solutions used for oral rehydration of infants and children with acute diarrhea. We have used cholera toxin-treated rat small intestine as a model of secretory diarrhea to assess the efficacy of oral rehydration solutions by intestinal perfusion. All solutions tested reversed net water secretion but a hypotonic bicarbonate-free solution was more effective than other solutions, including the World Health Organization oral rehydration solution (p less than 0.003). Net sodium secretion persisted with all solutions tested but there was a significant linear relationship between sodium concentration of the solution perfused and net sodium transport (r = 0.75, p less than 0.05). Cholera toxin treatment alone and in combination with perfusion of oral rehydration solutions significantly reduced plasma sodium concentration and osmolality (p less than 0.05), the effects being most marked with low sodium solutions. Although direct parallelism between observations in this animal model of secretory diarrhea and human diarrheal disease has not been established as yet, the model may be useful in assessing clinical efficacy of new oral rehydration solutions and in systematic analysis of the relative benefits of their individual components.