Plasma concentrations of cortisol were measured in 9 subjects form 0800 h to 1800 h at 10-min intervals during resting periods and 5 min intervals during exercise and recovery. This was done to assess to effect of exercise on the patterns of plasma cortisol levels. At 1000 h, 90 min exercise at a moderate work level (55% of Vo2 max) produced a transient rise in plasma cortisol which averaged 11.9 pg/100 ml (SE plus or minus 1.2) and introduced subsequent suppression of meal-related increase, At a lower work load (25% of Vo2 max) a similar transient rise occurred, avering 11.7 pg/100 ml (SE plus or minus 2.2); latency rate of concentration change and magnitude and not significatntly different from those at a moderate work load. A strenous work level was required to produce a more rapid response with an increased secretion rate. When 90 min exercise at a moderate work levels was preformed at 1300 h. (i.e. coinciding with postprandial cortisol concentration peak), the increase in plasma cortisol concentration (1.3 pg/100ml (SE plus or minus 0.3) was significantly lower than that induced by the same exercise performed in the morning. These results demonstrate that high plasma levels of cortisol diminish the subsequent stress response and that exercise-induced and meal-induced increases are not active.