Megestrol acetate (MA) has glucocorticoid activity and can induce significant secondary adrenal suppression. We designed this study to determine the extent of adrenal insufficiency in cancer patients receiving MA by utilising a sensitive low-dose adrenocorticotropin (ACTH) stimulation test. Adrenal function was assessed by a low-dose (0.625 microg) ACTH (1-24) stimulation test in 30 patients receiving MA for metastatic cancer. 10 of the patients who failed this test underwent a standard (250 microg) test on another day. Adrenal function was also evaluated in 15 of the patients by measuring the excretion of free cortisol in 24-h urine samples. Peak serum cortisol levels following stimulation with low-dose (0.625 microg) ACTH (1-24) were <18 microg/dl in 16 of 30 (53%) patients, of whom 9 had a basal serum cortisol level of <5 microg/dl. Five of 16 poor responders to the low-dose test showed normal stimulation with the standard (250 microg) ACTH (1-24) test. Thus, adrenal insufficiency would fail to be detected by the standard high dose test in these patients. Patients who failed the low-dose ACTH (1-24) test had lower 24-h urinary free cortisol excretion (8.7+/-10.3 microg/24 h) than normal responders (35+/-12.7 microg/24 h). Impaired adrenal function is common in cancer patients receiving MA. The low-dose ACTH (1-24) test is apparently capable of revealing adrenal insufficiency undetected by the standard high-dose ACTH test. Patients receiving MA might have inadequate adrenal function during episodes of infection or after withdrawal of MA therapy and this may require prompt corticosteroid treatment.