Continuous 24-hour infusions of a maximally stimulating dose (1 microgram/kg/h) of corticotropin-releasing hormone (CRH) have been shown to cause elevations of plasma cortisol and ACTH, but the pattern of results were confounded by serum cortisol causing feedback changes. We have looked at ACTH responses to saline or CRH infusions over 24 h in 6 normal subjects who, in addition, received either placebo or metyrapone, an 11 beta-hydroxylase inhibitor which blocks the formation of cortisol and thus abolishes glucocorticoid feedback. Cortisol and ACTH levels were measured by radioimmunoassay. Before metyrapone, CRH infusion resulted in exaggerated ACTH peaks throughout the day, as compared with normal saline: there was no influence on the noctural rise in ACTH. Following metyrapone alone, absolute cortisol levels were lower but circadian rhythmicity was preserved. Circadian rhythm of ACTH was maintained, with a fall in the evening to 14.5 +/- 4 pg/ml (mean +/- SE) at midnight and an exaggerated rise overnight, reaching a peak level of 90 +/- 33 pg/ml at 07:00 h. Subjects receiving CRH with metyrapone showed a similar pattern of responses, but with further enhanced ACTH levels. The evening fall reached a nadir of 30 +/- 6 pg/ml at 01:00 h. With diminished glucocorticoid feedback the nocturnal rise in ACTH was augmented by CRH infusion, with a morning peak of 193 +/- 21 pg/ml at 07:00 h. Thus, continuous infusion of CRH in the absence of steroid feedback leads to a retention of the circadian rhythmicity in ACTH secretion, reset at a higher absolute level.(ABSTRACT TRUNCATED AT 250 WORDS)