One hundred and one postmenopausal patients with advanced breast cancer were enrolled in a randomized phase II clinical trial to investigate the clinical and hormonal response to aminoglutethimide administered at daily doses of 2 x 125 mg, 3 x 125 mg or 2 x 250 mg, with no addition of hydrocortisone. Among 71 evaluable patients 25% showed objective tumor response (three complete, 15 partial), at all three dose levels and irrespective of the major tumor site. Previous treatment with Tamoxifen had been successful in 75%. Out of the 18 responding patients 10 had estrogen receptor positive, four had estrogen receptor negative tumors; the receptor status was unknown in four other patients. Progression-free interval was more than 700 days in 50% of the responders. Drowsiness caused early drug withdrawal in one patient. Side-effects were very mild, comparing favorably with standard therapy of 250 mg aminoglutethimide q.i.d. plus hydrocortisone. Plasma estrogen levels were reduced by all doses to the same 50% or less as in patients on standard treatment. In nine out of 27 patients a further decrease of estrone levels could be monitored with clinically improved results in five. Plasma cortisol and mineralocorticoids remained normal throughout more than 6 months. The original role of hydrocortisone administration to suppress a reflex rise of ATH in 'medical adrenalectomy' with standard dose aminoglutethimide is no longer tenable. Further phase III comparative clinical results pending, low dose aminoglutethimide as an aromatase inhibitor may at present be considered as an appropriate second-line endocrine treatment with low toxicity and expense.