We examined the influence of various carbohydrates of fuel homeostasis and glycogen utilization during prolonged exercise. Seventy-five grams of glucose, fructose, or placebo were given orally to eight healthy males 45 min before ergometer exercise performed for 2 h at 55% of maximal aerobic power (VO2max). After glucose ingestion, the rises in plasma glucose (P less than 0.01) and insulin (P less than 0.001) were 2.4- and 5.8-fold greater than when fructose was consumed. After 30 min of exercise following glucose ingestion, the plasma glucose concentration had declined to a nadir of 3.9 +/- 0.3 mmol/l, and plasma insulin had returned to basal levels. The fall in plasma glucose was closely related to the preexercise glucose (r = 0.98, P less than 0.001) and insulin (r = 0.66, P less than 0.05) levels. The rate of endogenous glucose production and utilization rose similarly by 2.8-fold during exercise in fructose group and were 10-15% higher than in placebo group (P less than 0.05). Serum free fatty acid levels were 1.5- to 2-fold higher (P less than 0.01) after placebo than carbohydrate ingestion. Muscle glycogen concentration in the quadriceps femoris fell in all three groups by 60-65% (P less than 0.001) during exercise. These data indicate that fructose ingestion, though causing smaller perturbations in plasma glucose, insulin, and gastrointestinal polypeptide (GIP) levels than glucose ingestion, was no more effective than glucose or placebo in sparing glycogen during a long-term exercise.