In a randomized double-blind trial, infants with mild or moderate diarrhoeal dehydration were rehydrated orally either with a simple solution containing table sugar and salt (without potassium or bicarbonate) or with a complete glucose/electrolyte formula. All 32 given glucose/electrolyte solution and 27 (93%) of 29 infants given sugar/salt were successfully rehydrated with similar improvement in metabolic acidosis and rapidity of rehydration. The drawbacks to oral therapy with simple sugar/salt solution were the frequent development of hypokalaemia and greater volume of vomiting during treatment. Carefully prepared sugar/salt solution, if accompanied by adequate potassium supplementation, may be used as an alternative to the preferred glucose/electrolyte formula when the latter is unavailable.