To investigate the mechanism underlying disturbances in hypothalamopituitary-adrenal (HPA) function in depressed patients, the dexamethasone suppression test (DST) was compared with a cortisol suppression test (CST) and placebo treatment in depressed patients and control subjects. Plasma levels of cortisol, ACTH and beta-endorphin were assessed at 3 times during the day after treatment with a single dose of exogenous steroid. Both dexamethasone and cortisol treatment resulted in suppression of cortisol, ACTH and beta-endorphin in control subjects, while neither treatment had any effect on the hormone levels in those depressed patients who showed cortisol nonsuppression after dexamethasone. In the depressed patients who were cortisol suppressors after dexamethasone, cortisol treatment only slightly changed plasma levels of beta-endorphin, although they were suppressed after dexamethasone treatment. In addition, high levels of both cortisol and beta-endorphin were observed after placebo treatment in all depressed patients compared to control subjects, probably due to the absence of the normally occurring decrease of these hormones during the day in these patients. Cortisol treatment, but not dexamethasone treatment, discriminated depressed patients from controls with respect to their beta-endorphin plasma levels. However, it is not yet clear whether these different effects of the two steroids are related to a different mode of action of these steroids in depressed patients. beta-Endorphin seems to be a useful marker in detecting disturbances in HPA function among depressed patients.