One of the main targets of fluid therapy in premature infants is to avoid variations in osmolality, which mainly means providing a stable sodium, glucose, and acid-base balance. Water, sodium, and acid-base balance were measured in 20 infants appropriate-for-gestational age with a gestational age less than or equal to 34 weeks. The infants were randomly assigned to one of two treatment groups. Fluid intake was restricted and air humidity in the incubator was high in order to minimize insensible water loss. Sodium intake in Group 1 was 2 mmol/kg/day and consisted of sodium chloride. Sodium intake in Group 2 was 4 mmol/kg/day and consisted of both sodium chloride and acetate. Weight loss was appropriate in both groups. In the high sodium intake group there was a tendency towards a more stable plasma sodium concentration than in the low sodium intake group. The use of sodium acetate was efficient and practical as normal acid-base balance was maintained. The protocol with restricted fluid intake (1st day 50 ml/kg, 2nd day 70 ml/kg, 3rd day 90 ml/kg, and 4th day 110 ml/kg), high air humidity, a sodium supply of 3 to 4 mmol/kg/day, and a slow correction of metabolic acidosis with sodium acetate, yields suitable guidelines in planning fluid and electrolyte therapy in premature infants less than or equal to 34 weeks' gestation.