Ninety postmenopausal women with advanced breast cancer were randomly assigned to be treated with HD-MPA administered either by oral route (daily dose 900 mg) or by intramuscular injections (1 g IM daily X 5 q w during 4 consecutive weeks followed by maintenance with 1 g twice weekly). Among 78 evaluable cases, most heavily pretreated, remissions, lasting for a median duration of 11 months, were more frequent on oral (8/37 = 22%) than on IM therapy (5/41 = 12%). In both arms, high estrogen receptor levels and various clinical factors were associated with higher response rates i.e., age greater than 60, Karnofsky greater than 70, light prior systemic treatment. Side-effects, consisting mainly of weight gain, hypertension and tremor occurred with equal frequency on oral or IM treatment. Five patients complained of pain at the sites of IM injections. Thus, we recommended that, whenever possible, the oral route should be preferred. During the same study, in 20 patients (11 on oral and 9 on IM therapy), blood was drawn at 0, 30, and 60 days of treatment for the assessment of MPA and hormone levels. In both arms, at 60 days, comparable levels of circulating MPA were obtained, with a very significant drop of cortisol, androstenedione, and estrone. These endocrine results, together with our clinical data, indicate that HD-MPA therapy is active on estrogen-dependent tumors with the same specificity as that of other modalities aiming to suppress the adrenal function. Its antineoplastic action in humans could be ascribed at least in part to its suppressive action on the adrenals, resulting in a severe estrogenic deprivation in postmenopausal women.