Oral rehydration among infants aged 0-3 months has not been adequately investigated. A controlled, randomized study was thus conducted in 65 young infants hospitalized with acute noncholera dehydrating diarrhea. The study was designed to compare the efficacy and safety of the standard WHO oral glucose-electrolyte solution containing 90 mmol of sodium per liter (Group A: 22 infants) with that of an oral glucose-electrolyte solution containing 60 mmol of sodium per liter (Group B: 22 infants) and with standard intravenous therapy (Group C: 21 infants). Among the 44 infants in Groups A and B, none required intravenous therapy. Dehydration, acidosis, and initial hyponatremia or hypokalemia were corrected with equal efficacy in all the three groups. In the critical first 8 h, the mean sodium absorption was significantly higher (p less than 0.01) in Group A. This resulted in hypernatremia (50%), periorbital edema (50%), mild pedal edema (27%), excessive irritability, and convulsions (4.5%). The mean serum sodium levels at 8, 24, and even 48 h were significantly higher (p less than 0.05) than those in Groups B and C. It is concluded that glucose-electrolyte oral solution containing 60 mmol of sodium per liter is as safe and effective as intravenous rehydration for the treatment of noncholera neonatal and early infantile diarrhea, while the standard WHO solution carries a significant risk of hypernatremia under similar conditions.