A total of 523 postmenopausal breast cancer patients with progressive disease were entered in a radomized, double-blind study of four dosages of diethylstibestrol (DES): 1.5, 15, 150, OR 1,500 MG/DAY. Higher dosages produced significantly (p less than .05) higher regression rates: 21% for the 1,500 mg dosage, 17% for the 150 mg dosage, 15% for the 15 mg dosage, and 10% for 1.5 mg dosage. Durations of regressions were similar regardless of the dosages used to induce them. Although the highest dosage produced the highest regression rate overall, selecting the best dosage or treatment of choice for each type of patient based on menopausal age and on dominant metastatic site would result in more regressions.