The integrated concentration (I.C.) of plasma cortisol was determined every 20 min during a 24-hr period, using a blood collection system composed of a non-thrombogenic catheter and a small, portable withdrawal pump. Thirteen experiments were undertaken in 11 adult subjects under daily routine life while a trained observer recorded physical activity as well as psychological reactions. The 24-hr studies were divided in six periods: (I) first 4 hr of sleep; (II) remainder of sleep; (III) from arousal to noon; (IV) from noon to 4 PM; (V) from 4 to 8 PM; (VI) from 8 PM to beginning of sleep. The following was demonstrated: a) The 24-hr profile of plasma cortisol is characterized by consistently high values during periods II and III. The mean I.C. of periods I and VI were similar and significantly lower than those of the other four periods, b) The introduction of the indwelling catheter had no significant influence on the I.C. of cortisol, nor did the food intake, c) It appeared that some episodes of increased secretory rate during the waking hours were closely related to stressful situations, d) The half-life of cortisol calculated on the basis of 63 episodes of 4 or more I.C. values showing a downtrend was 93 min. The finding of a large number of downfall episodes with prolonged half-life suggests that complete secretory inactivity is a rare event, e) The mean values of the 20-min I.C. of plasma cortisol for all subjects resulted in a curve of diurnal variation reminiscent of those reported previously when blood samples were collected at various intervals of time. The overall mean (±SD) value for 24-hr I.C. for the 11 subjects was 6.3 ± 0.9 μg/100 ml plasma.