Nine males cycled at 53% (s = 2) of their peak oxygen uptake (VO(2peak)) for 90 min (dry bulb temperature: 25.4 degrees C, s = 0.2; relative humidity: 61%, s = 3). One litre of flavoured water at 10 (cold), 37 (warm) or 50 degrees C (hot) was ingested 30 - 40 min into exercise. Immediately after the 90 min of exercise, participants cycled at 95%VO(2peak) to exhaustion to assess exercise capacity. Rectal and mean skin temperatures and heart rate were recorded. The gradient of rise in rectal temperature was influenced (P < 0.01) by drink temperature. Mean skin temperature was highest in the hot trial (cold trial: 34.2 degrees C, s = 0.5; warm trial: 34.4 degrees C, s = 0.5; hot trial: 34.7 degrees C, s = 0.6; P < 0.01). Significant differences were observed in heart rate (cold trial: 132 beats . min(-1), s = 13; warm trial: 134 beats . min(-1), s = 12; hot trial: 139 beats . min(-1), s = 13; P < 0.05). Exercise capacity was similar between trials (cold trial: 234 s, s = 69; warm trial: 214 s, s = 52; hot trial: 203 s, s = 53; P = 0.562). The heat load and debt induced via drinking resulted in appropriate thermoregulatory reflexes during exercise leading to an observed heat content difference of only 33 kJ instead of the predicted 167 kJ between the cold and hot trials. These results suggest that there may be a role for drink temperature in influencing thermoregulation during exercise.