Substrate utilization after fructose, glucose, or water ingestion was examined in four male and four female subjects during three treadmill runs at approximately 75% of maximal O2 uptake. Each test was preceded by three days of a carbohydrate-rich diet. The runs were 30 min long and were spaced at least 1 wk apart. Exercise began 45 min after ingestion of 300 ml of randomly assigned 75 g fructose (F), 75 g glucose (G), or control (C). Muscle glycogen depletion determined by pre- and postexercise biopsies (gastrocnemius muscle) was significantly (P less than 0.05) less during the F trial than during C or G. Venous blood samples revealed a significant increase in serum glucose (P less than 0.05) and insulin (P less than 0.01) within 45 min after the G drink, followed by a decrease (P less than 0.05) in serum glucose during the first 15 min of exercise, changes not observed in the C or F trials. Respiratory exchange ratio was higher (P less than 0.05) during the G than C or F trials for the first 5 min of exercise and lower (P less than 0.05) during the C trial compared with G or F for the last 15 min of exercise. These data suggest that fructose ingested before 30 min of submaximal exercise maintains stable blood glucose and insulin concentrations, which may lead to the observed sparing of muscle glycogen.