To examine the functioning of the hypothalamo-pituitary-adrenocortical axis in secondary adrenocortical insufficiency, we administered 100 micrograms synthetic human CRH, iv, plus 10 U lysine-8-vasopressin (LVP), im, three times daily for 3 consecutive days. The changes in plasma ACTH and cortisol levels during the administration and the response to an insulin tolerance test (ITT) conducted before and after the administration were determined. In three patients with isolated ACTH deficiency, basal plasma ACTH and cortisol levels were undetectablly low, and there was no response noted in the ITT or during CRH-LVP administration throughout the observation period. In four patients with adrenocortical insufficiency who had undergone successful transsphenoidal microadenomectomy for Cushing's disease and in six patients who had undergone curative unilateral adrenalectomy for Cushing's syndrome, basal plasma ACTH levels were low, but responded considerably to both stimulation tests. Along with the 3 days of CRH-LVP stimulation, however, neither the peak nor the time-integrated ACTH response was significantly enhanced, because of the variability of the responses among the patients. Compared with the ACTH response on the last day of CRH-LVP stimulation, the subsequent ITT tended to induce a lower ACTH response in the post-Cushing's disease patients and a higher response in the post-Cushing's syndrome patients. Regarding the plasma cortisol levels, the basal, peak, and integrated responses tended to increase daily during CRH-LVP administration. Conversely, the ITT after repetitive CRH-LVP administration induced a higher cortisol response than the test before CRH-LVP administration in the post-Cushing's disease patients. No serious complications were noted in any of the patients during or after the treatment. The present findings indicate that 1) repetitive administration of CRH in combination with LVP is a safe and valuable provocation test to examine the pituitary ACTH reserve and the integrity of the pituitary-adrenocortical axis; 2) isolated ACTH deficiency is usually due to a defect at the pituitary level; 3) with respect to adrenocortical responsiveness, post-Cushing's disease patients show a better accumulation of the provocative effect than do post-Cushing's syndrome patients; and 4) both hypothalamic and pituitary dysfunction are responsible for adrenal hypofunction in patients after hypercortisolemia, but post-Cushing's syndrome patients (especially those with a short period of hypercortisolemia) appeared to have less impairment of hypothalamic ACTH-releasing activity than post-Cushing's disease patients.