Perinatal glucose, insulin, and sodium homeostases were assessed in relation to antepartum intravenous infusions administered to 59 normal mothers undergoing cesarean section at term without labor under epidural anesthesia. Group A (N = 20) received 1 L of Ringer's lactate without dextrose during one hour; group B (N = 20), 1 L of 5% dextrose in water during one hour; and group C (N = 19), 1 L of 5% dextrose in water during two and one half hours. Mean maternal and fetal serum glucose and insulin and sodium concentrations at delivery differed among all groups in direct relationship to the rate of glucose infusion. Neonatal hypoglycemia (30 mg/dL or less) correlated with the presence of a glucose infusion, a maternal glucose concentration of 117 mg/dL or greater, and an umbilical venous insulin concentration of 26 microU/mL or greater. Among group A patients who received sodium, and group B and C patients who did not, fetal hyponatremia (umbilical venous sodium 130 mEq/L or less) correlated with the absence of sodium in the prepartum infusion. The results suggest that the antepartum administration of a balanced electrolyte solution without excess glucose infusion can minimize the incidence of fetal hyperglycemia and hyponatremia and neonatal hypoglycemia.