To investigate the mechanisms responsible for glucocorticoid feedback on nonstress-induced ACTH secretion in normal subjects, 24 volunteer subjects (14 males and 10 females, 21 to 43 years) were divided into six groups in a randomized fashion and studied. Each subject received a single midnight dose of 30 mg/kg per body weight of metyrapone and then cortisol was administered according to different protocols in the next morning to provide extreme variations of the input signal. It was found that no obvious inhibition in plasma ACTH levels was shown during the first 15 min despite the fact that cortisol was given at rather larger doses and short time intervals. However, a significant suppression in plasma ACTH levels began to manifest approximately 30 min after cortisol administration in each study group and it became apparent that the degree of inhibition of ACTH level at 75 min correlated with the plasma cortisol concentrations at the same moment (r = 0.97, P less than 0.01) as well as with the dosage of cortisol during this time, whatever administered (r = 0.99, P less than 0.01). In summary, our data provided evidence for a delayed, proportional, and integral phase of glucocorticoid feedback on nonstress-induced ACTH secretion in normal human volunteers.