Twelve 9- to 12-year-old children (6 boys, 6 girls) performed four exercise-in-heat (35 degrees C, 45% RH) trials which differed in the composition of the fluids they drank. In each trial, subjects cycled for one 20-min and two 15-min bouts at 50% peak VO2 with 10-min rest periods in between. In a fourth bout, they cycled at 90% peak VO2 until exhaustion. Drinks had the same grape flavor and were assigned in a double-blind design and in a Latin-square order. Subjects drank 7 ml.kg-1.h-1 to keep them euhydrated. Three of the drinks had 6% carbohydrates (CHO), with different [Na+]: 0, 8.8, 18.5 mmol.l-1 and one drink had neither CHO nor Na+ (WATER). Among drink trials, there were no differences in the increase in rectal temperature, HR, or performance time to exhaustion. Despite the larger Na+ deficit induced by the Na(+)-free drinks compared with the Na+ drinks (11.8 +/- 1.4 vs 5.7 +/- 0.9 mmol.h-1), neither plasma [Na+] nor osmolality were affected. These results suggest that electrolyte, as in the above conditions, did not affect electrolyte balance, thermoregulatory responses, or aerobic performance of children exercising in the heat. The greater Na+ deficit induced by ion-free drinks was of minor biological importance.