To examine the effects of a 3-day high carbohydrate (H-CHO) and low carbohydrate (L-CHO) diet on 45 min of cycling exercise, 12 endurance-trained cyclists performed a 45-min cycling exercise at 82 +/- 2% VO2peak following an overnight fast, after a 6-day diet and exercise control. The 7-day protocol was repeated under 2 randomly assigned dietary trials H-CHO and L-CHO. On days 1-3, subjects consumed a mixed diet for both trials and for days 4-6 consumed isocaloric diets that contained either 600 g or 100 g of carbohydrates, for the H-CHO and the L-CHO trials, respectively. Muscle biopsy samples, taken from the vastus lateralis prior to the beginning of the 45-min cycling test, indicated that muscle glycogen levels were significantly higher (p < .05) for the H-CHO trial (104.5 +/- 9.4 mmol/kg wet wt) when compared to the L-CHO trial (72.2 +/- 5.6 mmol/kg wet wt). Heart rate, ratings of perceived exertion, oxygen uptake, and respiratory quotient during exercise were not significantly different between the 2 trials. Serum glucose during exercise for the H-CHO trial significantly increased (p < .05) from 4.5 +/- 0.1 mmol x L(-1) (pre) to 6.7 +/- 0.6 mmol x L(-1) (post), while no changes were found for the L-CHO trial. In addition, post-exercise serum glucose was significantly greater (p < .05) for the H-CHO trial when compared to the L-CHO trial (H-CHO, 6.7 +/- 0.6 mmol x L(-1); L-CHO, 5.2 +/- 0.2 mmol x L(-1)). No significant changes were observed in serum free fatty acid, triglycerides, or insulin concentration in either trial. The findings suggest that L-CHO had no major effect on 45-min cycling exercise that was not observed with H-CHO when the total energy intake was adequate.