OBJECTIVE This study aimed to examine the effect of glycemic index of pre-exercise carbohydrate (CHO) ingestion on exercise metabolism and performance. METHODS Eight endurance trained men ingested a high glycemic index (HGI), low glycemic index (LGI), or a placebo (CON) meal 45 min before exercise and then cycled for 50 min at 67% VO2max. Subjects subsequently performed a 15-min self-paced performance ride in which total work (kJ) was recorded. RESULTS Plasma glucose concentrations were higher (P < 0.01) after ingestion in HGI compared with LGI and CON (7.53 +/- 0.64 vs 5.55 +/- 0.21 and 4.65 +/- 0.14 mmol.L-1 for HGI, LGI, and CON, respectively, 30 min postprandial; mean +/- SE) but declined at the onset of exercise and were lower (P < 0.01) compared with LGI and CON (4.03 +/- 0.31 vs 4.64 +/- 0.24 and 5.09 +/- 0.16 mmol.L-1 for HGI, LGI, and CON respectively; mean +/- SE) at 10 min of exercise. Plasma glucose remained depressed (P < 0.01) until 30 min into exercise in HGI compared with other trials. Plasma insulin concentrations were higher (P < 0.01) following ingestion during rest and exercise in HGI compared with LGI and CON. Plasma FFA concentrations were lower (P < 0.05) following ingestion in HGI and LGI compared with CON and higher (P < 0.05) in LGI compared with HGI at the start and end of exercise. RER and CHO oxidation was higher (P < 0.01) in HGI compared with LGI and CON during submaximal exercise. There were no differences in work output during the performance cycle. CONCLUSIONS These data indicate that pre-exercise CHO feedings with varying glycemic indexes do not affect exercise performance following short term submaximal exercise despite alterations in metabolism.