Nine male student games players consumed either flavoured water (0.1 g carbohydrate, Na+ 6 mmol x l(-1)), a solution containing 6.5% carbohydrate-electrolytes (6.5 g carbohydrate, Na+ 21 mmol x l(-1)) or a taste placebo (Na+ 2 mmol x l(-1)) during an intermittent shuttle test performed on three separate occasions at an ambient temperature of 30 degrees C (dry bulb). The test involved five 15-min sets of repeated cycles of walking and variable speed running, each separated by a 4-min rest (part A of the test), followed by 60 s run/60 s rest until exhaustion (part B of the test). The participants drank 6.5 ml x kg(-1) of fluid as a bolus just before exercise and thereafter 4.5 ml x kg(-1) during every exercise set and rest period (19 min). There was a trial order effect. The total distance completed by the participants was greater in trial 3 (8441 +/- 873 m) than in trial 1 (6839 +/- 512, P < 0.05). This represented a 19% improvement in exercise capacity. However, the trials were performed in a random counterbalanced order and the participants completed 8634 +/- 653 m, 7786 +/- 741 m and 7099 +/- 647 m in the flavoured water (FW), placebo (P) and carbohydrate-electrolyte (CE) trials, respectively (P = 0.08). Sprint performance was not different between the trials but was impaired over time (FW vs P vs CE: set 1, 2.41 +/- 0.02 vs 2.39 +/- 0.03 vs 2.39 +/- 0.03 s; end set, 2.46 +/- 0.03 vs 2.47 +/- 0.03 vs 2.47 +/- 0.02 s; main effect time, P < 0.01). The rate of rise in rectal temperature was greater in the carbohydrate-electrolyte trial (rise in rectal temperature/duration of trial, degrees C x h(-1); FW vs CE, P < 0.05; P vs CE, N.S.). Blood glucose concentrations were higher in the carbohydrate-electrolyte than in the other two trials (FW vs P vs CE:rest, 4.4 +/- 0.1 vs 4.3 +/- 0.1 vs 4.2 +/- 0.1 mmol x l(-1); end of exercise, 5.4 +/- 0.3 vs 6.4 +/- 0.6 vs 7.2 +/- 0.5 mmol x l(-1); main effect trial, P < 0.05; main effect time, P < 0.01). Plasma free fatty acid concentrations at the end of exercise were lower in the carbohydrate-electrolyte trial than in the other two trials (FW vs P vs CE: 0.57 +/- 0.08 vs 0.53 +/- 0.11 vs 0.29 +/- 0.04 mmol x l(-1); interaction, P < 0.01). The correlation between the rate of rise in rectal temperature (degrees C x h(-1)) and the distance completed was -0.91, -0.92 and -0.96 in the flavoured water, placebo and carbohydrate-electrolyte conditions, respectively (P < 0.01). Heart rate, blood pressure, plasma ammonia, blood lactate, plasma volume and rate of perceived exertion were not different between the three fluid trials. Although drinking the carbohydrate-electrolyte solution induced greater metabolic changes than the flavoured water and placebo solutions, it is unlikely that in these unacclimated males carbohydrate availability was a limiting factor in the performance of intermittent running in hot environmental conditions.