Effects of carbohydrate (CHO) supplementation on muscle glycogen utilization and endurance were evaluated in seven well-trained male cyclists during continuous cycling exercise that varied between low [45% maximal O2 uptake (VO2 max)] and moderate intensity (75% VO2 max). During each exercise bout the subjects received either artificially flavored placebo (P), 10% liquid CHO supplement (L; 3 x 18 g CHO/h), or solid CHO supplement (S; 2 x 25 g CHO/h). Muscle biopsies were taken from vastus lateralis during P and L trials immediately before exercise and after first (124 min) and second set (190 min) of intervals. Subjects then rode to fatigue at 80% VO2 max. Plasma glucose and insulin responses during L treatment reached levels of 6.7 +/- 0.7 mM and 70.6 +/- 17.2 microU/ml, respectively, and were significantly greater than those of P treatment (4.4 +/- 0.1 mM and 17.7 +/- 1.6 microU/ml) throughout the exercise bout. Plasma glucose and insulin responses of S treatment were intermediate to those of L and P treatments. Times to fatigue for S (223.9 +/- 3.5 min) and L (233.4 +/- 7.5 min) treatments did not differ but were significantly greater than that of P treatment (202.4 +/- 9.8 min). After the first 190 min of exercise, muscle glycogen was significantly greater during L (79 +/- 3.5 mumol/g wet wt) than during P treatment (58.5 +/- 7.2 mumol/g wet wt). Furthermore, differences in muscle glycogen concentrations between L and P treatments after 190 min of exercise and in time to fatigue for these treatments were positively related (r = 0.76, P < 0.05). These results suggest that CHO supplementation can enhance prolonged continuous variable-intensity exercise by reducing dependency on muscle glycogen as a fuel source.