The association of hyperglycemia and markers of hepatic dysfunction with dextrose infusion rates in Korean patients receiving total parenteral nutrition (TPN) was studied. A retrospective study of 122 patients with normal glucose levels and liver function tests (LFTs) was conducted. Pharmacy and medical records of all patients who received TPN from three university-affiliated teaching hospitals in Korea between January 1998 and December 1999 were reviewed. Each patient was categorized as receiving dextrose at (1) < or = 5 or > 5 mg/kg/min and (2) < or = 4, 4.1-5, 5.1-6, or > 6 mg/kg/min. Fifty-five patients received dextrose at a rate of > 5 mg/kg/min for 15.1 +/- 12.8 days and 67 patients at a rate of < or = 5 mg/kg/min for 10.1 +/- 6.8 days. Two patients in each group did not have follow-up glucose levels. Of the 53 patients in the > 5 mg/kg/min group, 16 exhibited hyperglycemia, compared with 21 of the 65 receiving lower rates of dextrose infusion. Elevated aspartate transaminase was the most common abnormal LFT value in both groups (25% and 29% in the < or = 5- and > 5-mg/kg/min groups, respectively). In the group receiving dextrose at > 5 mg/kg/min, 22.2% had two hepatic enzyme levels elevated concurrently, while 18.5% had two hepatic enzyme levels elevated in the group receiving dextrose at < or = 5 mg/kg/min. Regression analysis revealed that duration of TPN and dextrose infusion rate were positively correlated with blood glucose levels and that duration of TPN was positively correlated with abnormal LFT values. A retrospective study of Korean patients revealed no significant difference in the risk of hyperglycemia or hepatic dysfunction between those receiving < or = 5 and > 5 mg/kg/min dextrose infusion in their TPN.