The effects of a 2-h infusion of a low dose of cortisol on concentrations of adrenocorticotropic hormone (ACTH) and cortisol were studied in six inpatients with recurrent major depression and six healthy volunteers. Each subject was studied twice and received, in random order, from 11:00 to 13:00 h a 25 ml/h infusion of either 3 mg/h of cortisol or saline. Blood samples for ACTH and cortisol determination were drawn between 10:45 and 13:00 h every 15 min. ACTH and cortisol measurements in patients did not differ significantly from those in volunteers at any of the time points tested. The finding of an intact intermediate feedback in depression, where nonsuppression on the dexamethasone suppression test is frequently observed, may be explained by the binding of cortisol at limbic and hypothalamic corticosteroid receptors, while dexamethasone acts primarily at the pituitary. Findings of this pilot investigation should be confirmed in larger groups of patients for whom data from the dexamethasone suppression test are also available.