It has been suggested that the well-documented hypercortisolaemia found in a proportion of patients with severe depression occurs either in response to excessive secretion of corticotrophin-releasing hormone-41 (CRH-41) from the hypothalamus, or as a consequence of up-regulation of pituitary CRH-41 receptors. The attenuation of the normal ACTH response to CRH-41 in these subjects is thought to result from inhibition of corticotrophin secretion by elevated cortisol levels. We tested these hypotheses by examining ACTH responses to metyrapone, an 11β-hydroxylase inhibitor which blocks the formation of cortisol, followed by CRH-41 in 15 severely depressed in-patients diagnosed according to DSM-IIIR criteria. Patients were assigned to two groups according to their response to overnight administration of 1 mg dexamethasone: suppressors (8) and nonsuppressors (7). A third group consisted of 6 healthy matched controls. Metyrapone 750 mg was given 4-hourly for 24 h and samples were taken for cortisol and ACTH. Six of the original 15 patients (3 from each group) were given a bolus dose of 100 µg human CRH-41 intravenously after 24 h of metyrapone, and ACTH levels were measured over 2 h. Falls in circulating cortisol in response to metyrapone were similar in all three groups. However, we found exaggerated rises in ACTH amongst the nonsuppressors, as compared to the suppressors and the control group, after metyrapone. At 24 h the mean ACTH amongst the nonsuppressors was 134 ± 43 pg/ml(mean ± SD), a value significantly greater (p < 0.01) compared to 49 ± 11 pg/ml among the suppressors and 57 ± 9 pg/ml amongst the controls. After CRH-41, suppressors showed a greater rise in ACTH compared to nonsuppressors, but the proportional responses were similar. Our findings support the hypothesis that the hypercortisolaemia and dexamethasone resistance seen in a proportion of patients with depression are a consequence of increased CRH-41 secretion.