Eighteen children were infused with 10% solutions of glucose, invert sugar and fructose during a period of 6 hours at a rate of infusion of 1.25 gm of sugar per kilogram of body weight per hour. All but two of the children had less glycosuria from intravenous administration of glucose than from the administration of invert sugar or fructose. In 16 children glucose appeared in the urine only during the first 2 or 3 hours of infusion of glucose, but fructose was generally present in the urine during almost all of the 6 hours of infusion of invert sugar or of fructose. Eleven of the eighteen patients had more melituria during the first 2 hours of infusion of glucose than during the first 2 hours of infusion of invert sugar or fructose. This contrasts with the total amount excreted over an 8-hour period. When glucose was administered, total concentration of sugar in the blood obtained during the first 2 hours of infusions were higher than concentrations obtained during infusions of invert sugar and fructose; subsequently, there was often little difference in total concentration of sugar in the blood during infusion of the three hexoses. Determinations of concentrations of sugars in the blood revealed that the renal threshold for glucose was above 175 mg/ 100 ml of blood, while the threshold for fructose was frequently as low as 10 mg/ 100 ml of blood. There was no significant difference in total volumes of urine obtained from each patient at the end of 8-hour collection periods after infusions of the three carbohydrates. The present study suggests that in some children who require intravenous fluid therapy with maximum caloric intake for periods longer than 2 hours glucose is to be preferred to invert sugar or fructose. Whether glucose is the best choice for children after operation or in situations of severe stress remains to be demonstrated.