The responses of six healthy male subjects to submaximal and maximal exercise on a stationary bicycle ergometer have been investigated over a 24-hour period. Measurements were made on each subject at approximately three-hourly intervals and they included minute ventilation at a carbon dioxide output of 1-5 1 min-minus 1 (VE 1-5), tidal volume at a fixed VE of 30 1 min-minus 1 (VT 30), oxygen intake (VO2) at a work load (W) of 150 W (VO2 150), tympanic temperature (Tty) and cardiac frequency at a VO2 of 1-5 1 min-minus 1 (fH 1-5). The experiments were conducted in three parts: on the first occasion two subjects were measured during exercise; on the second occasion a further four subjects were observed in a similar way but starting from a baseline of zero load, and the measurements also included an estimate of cardiac output (Q) using a rebreathing technique. Finally the maximum aerobic power output (VO2max) was measured in three of the subjects in early morning and late evening. Diet and habitual physical activity were held constant between the exercise test on all three occasions. The results show that in the first two subjects fH 1-5 and Tty had a rhythmic pattern of variation with time of day whereas VE 1-5, VT30, and VO2 150 remained fairly constant. The variation in fH 1-5 was associated with Tty; the two variables reached a minimum at similar to 0500 hr and a maximum at similar 1200 hr. These results were confirmed on the remaining subjects but the changes in fH 1-5 and Tty were shown to be more variable and reduced in magnitude. Further, if the changes were calculated from a baseline of zero load, it was shown that the absolute changes observed in fH 1-5 and Tty were not due to the exercise per se but to changes in the basal level from which each subject operated. In addition it was shown that VO2 max and Q remained constant and were independent of the time of day. It is concluded that provided the exercise test conditions are rigidly standardized and subjects exercise from a controlled baseline there is no evidence for circadian variation in the change of responses to work at submaximal or maximal effort.