The possibility of assessing hypothalamic-pituitary-adrenal (HPA) function by the standard ACTH test (250 microg) has been widely discussed in the past years and compared with the role of the insulin tolerance test (ITT). Recently, it was shown that low doses of ACTH, such as 1 microg i.v., induce a maximal adrenal response and, by reducing the discrepancies compared with the ITT also allow one to detect mild forms of secondary hypoadrenalism. In the present study the 1 microg ACTH test was performed in patients with hypothalamic-pituitary disease in order to assess adrenal function, and the results have been compared with those obtained after the insulin test. Fifty-seven patients (31 men and 26 women, aged 19-73 years) with hypothalamic-pituitary diseases were studied: 51 patients were affected with pituitary tumor and 6 patients had hypothalamic disorders. All these patients and 18 healthy volunteers (7 men and 11 women, aged 19-46 years) received 1 microg i.v. ACTH injection. In addition, the ITT (0.1-0.15 U/kg body weight) was performed in all patients. In normal subjects mean cortisol levels significantly (P<0.001) increased from a baseline of 393+/-43 nmol/l to a peak of 770+/-41 nmol/l after 1 microg ACTH. In 44 patients with hypothalamic-pituitary disease 1 microg ACTH caused a cortisol rise similar to that of normal subjects (from 332+/-17 to 769+/-24 nmol/l; P<0.001), while an impaired response (from 124+/-23 to 312+/-46 nmol/l) was observed in 13 cases (23%), 7 of them with low morning cortisol levels (10-127 nmol/l) and 6 with basal values at the lower limit of normality. The cortisol response to ITT was compared with that obtained after the 1 microg ACTH test: 10 patients failed both challenges, 4 patients who passed the ACTH test failed the ITT, while 3 patients who failed the ACTH test passed the ITT. The 23 out of 57 patients (40%) who showed a cortisol peak greater than 750 nmol/l after 1 microg ACTH had a normal response to ITT. A positive correlation between cortisol peaks after ACTH and after insulin was also found (r = 0.68, P<0.001). Assuming a 100% accuracy of ITT, the low dose ACTH test yielded a 71% sensitivity and a 93% specificity. In conclusion, the low-dose ACTH test is a useful, safe and inexpensive tool for the initial assessment of HPA function in patients with hypothalamic-pituitary disease. In fact, the ITT is unnecessary when cortisol peaks are greater than 750 nmol/l after 1 microg ACTH and also when very low cortisol basal levels indicate an overt hypoadrenalism. Within these limits the ITT is mandatory and its important role in the recognition of secondary adrenal failure is further confirmed.