Major surgery is associated with fall in the concentration of inorganic phosphate in serum, as is intravenous infusion of glucose. Hypophosphatemia during different forms of postoperative dextrose administration was evaluated in patients who had undergone colorectal surgery. They were randomized to two groups. All patients received standardized intravenous fluids on the first 3 postoperative days, but one group had constant infusion of a solution containing glucose (4%), sodium (40 mmol/l) and potassium (20 mmol/l) throughout the observation period, while the other group had a 5-hour infusion of 10% glucose daily, with potassium and sodium solution in between. The amounts of administered glucose and electrolytes were the same in both groups. The serum phosphate levels were significantly lower in the group with constant glucose infusion, due to intergroup difference in renal handling of phosphate. Significantly less phosphate was reabsorbed in the proximal tubules when glucose was given as 24-hour infusion than in the group with 5-hour infusion.