Two groups of 12 cholecystectomized or vagotomized patients were given central venous infusions of either a combination of glucose, fructose and xylitol at the ratio of 1:2:1 or of glucose alone for 5 days. The dose was increased stepwise from 1.43 g/kg/h on operation day up to 7.14 g/kg/24 h on the fourth day after surgery, which equals 100-500 g/24 h for the average 70 kg patient. On both infusion regimens mean blood glucose values ranged from 98-124 mg/100 ml. Urinary losses of infused substrates amounted to 0.2-0.8 g/24 during glucose infusion and to 2.22-13.4 g/24 h during infusions of the carbohydrate combination. Regarding mean blood or serum values of lactate, pyruvate, uric acid, sodium, potassium, phosphorus, insulin and free fatty acids, no significant differences were found between the two types of carbohydrate infusion. Side effects were not observed. The results obtained allow conclusion that in clinical situations with moderate stress, such as the selective operations mentioned above, the combination of glucose, fructose and xylitol studied offers no advantage over glucose alone. However, in severely ill patients presenting with more pronounced glucose intolerance, further studies with sugar substitutes are warranted.