SUMMARY Concentration and specific activity of plasma cortisol were measured for 4 h after the intravenous injection of 10 μCi [1,2-3H]cortisol into 14 normal men. Of these subjects four were resting controls; four exercised at a high work load for 1 h; four exercised at a low work load for 1 h and two received infusions of unlabelled cortisol, all beginning 1 h after the administration of [1,2-3H]cortisol. In the high work load group plasma cortisol had increased by 12·2 ± 6·3 μg/100 ml at the 60th minute of exercise. In this group the half-life of [3H]cortisol was 31·5 ± 5·3 (s.d.) min in contrast to 74·5 ± 8·3 min for the resting controls (P < 0·001). In the light exercise group plasma cortisol concentration tended to fall, but due to large intersubject differences this was not statistically significant. In this group the half life (t½) of [3H]cortisol was 43·8 ± 3·9 min which was also significantly different from that of the resting controls (P < 0·001). Specific activity of plasma cortisol fell rapidly (mean t½ = 15 min) during 1 h of heavy exercise, and continued to fall to a nadir 10–30 min after exercise had ceased, finally reaching a value some 60% above the nadir 1·5–2·0 h after exercise had ceased. A similar, although exaggerated, pattern was observed in the two resting subjects into whom 5 and 4 mg respectively, of unlabelled cortisol were infused. In contrast, in the light exercise group cortisol specific activity changed only slowly (mean t½ = 151 min) but continued to fall after exercise. In the light exercise group, during the latter part of exercise and during the first hour after exercise, the ratio [3H]cortisone: [3H]cortisol in plasma was significantly higher than the corresponding values for the resting group (P < 0·05). Even higher values for this ratio were obtained for some of the heavy exercise subjects but due to wide scatter the group was not statistically significantly different from the resting group. These results suggest that exercise itself increases the rate of uptake of cortisol by peripheral tissues and that when the work load exceeds a critical level stimulation of the adrenal cortex results in a massive secretion of cortisol which is sufficient to raise the plasma level which in turn promotes further ingress of cortisol into the tissues. After exercise at a high work load a return of cortisol from the tissues to the plasma can be detected.