The main study comprised 16 patients undergoing colon surgery. On the day of operation and the 3 following days 100 g of glucose was infused at the rate of 0.3 g/kg/h. Half the patients had 10 mmol of phosphorus added to each 1 000 ml 10% glucose solution. The investigation demonstrated that two different kinds of hypophosphatemia occur in the immediate postoperative period. A significant decrease in fasting plasma phosphate was found at the first, second and the third postoperative morning, most pronounced at the second day (1.20 +/- 0.05 to 0.78 +/- 0.07 mmol/l). A significant correlation between these changes and the corresponding 24-hour phosphorus balance was demonstrated (r = 0.61, p < 0.001). The falls in fasting phosphate could not be prevented by phosphorus addition because an amount of phosphorus corresponding to the amount added was excreted in excess in the urine. The plasma phosphate was decreased furthermore during and even 4 hours after the 5-hour glucose infusion (from 0.76 +/- 0.05 to 0.49 +/- 0.07 nmol/l at the end of the infusion at the second day). This hypophosphatemia was prevented by the phosphorus addition.--In average 3% of the infused sugar was lost in the urine. The solitary examples of higher losses (10-20%) were not followed by a higher urinary production. It is therefore concluded that 0.3 g glucose/kg/h is a suitable infusion rate in the immediate postoperative period.