To study the effect of temperature on muscle metabolism during submaximal exercise, six endurance-trained men had one thigh warmed and the other cooled for 40 min prior to exercise using water-perfused cuffs. One cuff was perfused with water at 50-55¡C (HL) with the other being perfused with water at 0¡C (CL). With the cuffs still in position, subjects performed cycling exercise for 20 min at a work load corresponding to 70 % VO2,peak (where VO2,peak is peak pulmonary oxygen uptake) in comfortable ambient conditions (20-22¡C). Muscle biopsies were obtained prior to and following exercise and forearm venous blood was collected prior to and throughout the exercise period. Muscle temperature (Tmus) was not different prior to treatment, but treatment resulted in a large difference in pre-exercise Tmus (difference = 6á9 ± 0á9¡C; P < 0á01). Although this difference was reduced following exercise, it was nonetheless significant (difference = 0á4 ± 0á1¡C; P < 0á05). Intramuscular [ATP] was not affected by either exercise or muscle temperature. [Phosphocreatine] decreased (P < 0á01) and [creatine] increased (P < 0á01) with exercise but were not different when comparing HL with CL. Muscle lactate concentration was not different prior to treatment nor following exercise when comparing HL with CL. Muscle glycogen concentration was not different when comparing the trials before treatment, but the post-exercise value was lower (P < 0á05) in HL compared with CL. Thus, net muscle glycogen use was greater during exercise with heating (208 ± 23 vs. 118 ± 22 mmol kg-1 for HL and CL, respectively; P < 0á05). These data demonstrate that muscle glycogen use is augmented by increases in intramuscular temperature despite no differences in high energy phosphagen metabolism being observed when comparing treatments. This suggests that the increase in carbohydrate utilization occurred as a direct effect of an elevated muscle temperature and was not secondary to allosteric activation of enzymes mediated by a reduced ATP content.