Patients developing visceral breast cancer metastases generally receive chemotherapy rather than endocrine therapy. Recent aromatase inhibitor studies have reported activity in such patients; therefore, this study formally evaluated anastrozole and exemestane in postmenopausal patients in this setting.Postmenopausal women with advanced breast cancer and > or = 1 visceral (liver or lung) lesion were randomized to anastrozole (1 mg/day orally) or exemestane (25 mg/day orally) for > or = 8 weeks. The primary endpoint was objective response in visceral lesions based on modified Response Evaluation Criteria in Solid Tumors. Secondary endpoints included clinical benefit (objective response plus stable disease > or = 180 days), overall survival, and adverse events.A total of 130 patients were enrolled, and 128 patients (64 anastrozole, 64 exemestane) were included in the intent-to-treat analysis. Accrual delays caused study closure before the target enrollment (N = 200) was reached, limiting the statistical power of the study. Objective response in visceral sites was approximately 15% in both groups. Clinical benefit in visceral sites was 32% of the patients treated with anastrozole and 38% of the patients treated with exemestane. Median survival was 33.3 months and 30.5 months in the anastrozole and exemestane groups, respectively. Toxicities were similar to those previously reported; however, treatment-related adverse events were more frequent with anastrozole (41%) than with exemestane (31%). Both treatments were generally well tolerated in patients with postmenopausal breast cancer with visceral metastases.Efficacy was similar in both treatment groups for all endpoints. Aromatase inhibitors can be considered as a treatment option in postmenopausal patients with hormone receptor-positive visceral breast cancer metastases.