The response of plasma testosterone, LH and FSH to oral administration of cortisol or dexamethasone was studied for 26 h in 6 groups of 12 normal adult males. Twenty-four hour treatment with cortisol (60 mg at 1000 h, followed by 30 mg at 2 h intervals) or with dexamethasone (6 mg at 1000 h, followed by 3 mg at 6 h intervals) was without effect on testosterone during the day but suppressed the nocturnal rise completely. LH and FSH did not decrease at any time during the day and showed a highly significant increase during the second part of the night. Short-term treatment (1000-1600 h, 1600-2200 h or 2400 h-0800 h) with cortisol was less effective in suppressing the nocturnal rise in testosterone. It was concluded that administration of cortisol leading to plasma levels as seen under treatment with ACTH suppresses testosterone by abolishing or flattening the nocturnal rise. This effect was not mediated by changes in LH or FSH. Our data suggest that the ACTH-induced suppression of testosterone is due to an action of cortisol.