Nine healthy male subjects differing in their training status (VO2 max 54 +/- 7 ml.min-1.kg-1, mean +/- SD; 43-64 ml.min-1 kg-1, range) exercised on two occasions separated by one week. On each occasion, having fasted overnight, subjects exercised for 1 h on an electrically braked cycle ergometer at a workload equivalent to 70 per cent VO2 max (test A) or at a fixed workload of 140 W (test B). Each test was assigned in a randomized manner and was performed at an ambient temperature of 22.5 +/- 0.0 degrees C and a relative humidity of 85 +/- 0 per cent. Absolute exercise workload was the most successful predictor of sweat loss during test A (r = 0.82, p less than 0.01). Sweat loss was also related to VO2 max tests A (r = 0.67, p less than 0.05) and B (r = 0.67, p less than 0.05). There was no relationship between resting pre-exercise core temperature and VO2 max. However, core temperature recorded during the final min of exercise in test B was inversely related to VO2 max (r = -0.86, p less than 0.01). As a consequence, core temperature during the final minute of exercise was also related to the relative exercise intensity (% VO2 max) performed (r = 0.82, p less than 0.01). The heart rate response during test B was inversely related to VO2 max (r = -0.71, p less than 0.05) and was positively related to the relative exercise intensity performed (r = 0.68, p less than 0.05). No relationship was found between weighted mean skin temperature during the final minute of exercise and the relative (r = 0.26) or absolute (r = 0.03) workloads performed during exercise. The results of the present experiment suggest that cardiovascular fitness (as indicated by VO2 max) will have a significant influence upon the thermoregulatory responses of Man during exercise.