To assess the ability of desmopressin to differentiate between pituitary and ectopic ACTH-dependent Cushing's syndrome and to determine whether diagnostic accuracy could be improved by administering it together with human sequence CRH, we examined its effects on cortisol and ACTH secretion when given alone or in combination with CRH in patients with Cushing's syndrome of varied etiology and compared these data to the results of a standard CRH test in the same individuals. Each patient was studied on three occasions, in random order, separated by at least 48 h. At 0900 h, via an indwelling forearm cannula, 10 micrograms desmopressin, 100 micrograms CRH, or a combination of the two were given as an iv bolus; thereafter, blood was drawn every 15 min for 2 h. The responses to the individual agents were determined according to the timing and calculation criteria suggested by Nieman et al. (1993). A total of 25 patients with Cushing's syndrome were studied: 17 patients with pituitary-dependent Cushing's syndrome, Cushing's disease (CD); 5 patients with occult ectopic ACTH secretion (EC); and 3 patients with primary adrenal (ACTH-independent) Cushing's syndrome. In this series, the best discrimination among ACTH-dependent patient groups was achieved using the combined test. Using the responses of plasma cortisol, all 17 patients with CD showed a rise greater than any of the 5 patients with EC, whereas 1 patient with CD showed a plasma ACTH response within the range seen in the patients with EC. Plasma cortisol responses to desmopressin alone were seen in 14 of 17 patients with CD and 1 of 5 patients with EC and, after CRH alone, in 15 of 17 patients with CD but in no patient with EC. In contrast, plasma ACTH responses after CRH alone were seen in 14 of 17 patients with CD and 2 of 5 patients with EC and, after desmopressin alone, in 12 of 17 with CD and 3 of 5 with EC, thus indicating overlapping responses between the groups and poorer discrimination. No responses were seen in the ACTH-independent group. These data indicate that desmopressin causes the secretion of ACTH and cortisol in patients with ACTH-dependent Cushing's syndrome, and that in combination with CRH, it may provide an improvement over the standard CRH test in the differential diagnosis of ACTH-dependent Cushing's syndrome. Furthermore, these data suggest that there may be abnormalities in vasopressin receptor function or number in ACTH-secreting tumors.