OBJECTIVE To examine the outcomes of endocrine naive patients treated with tamoxifen as initial endocrine therapy for metastatic breast cancer. Data were obtained from the long-term follow-up of two previously published randomized trials. METHODS All patients received tamoxifen 20 mg po in a single daily dose. Eligibility required patients to be age > or = 18, performance status 0-3, and estrogen or progesterone receptor positive or unknown. Patients were ineligible if they had any prior endocrine therapy in either the adjuvant or metastatic setting. RESULTS 156 patients have been followed for a median of 8.3 years. Median age was 61 years, 83% were > or = 50 years, 84% performance status of 0-1, 43% were both ER and PR positive, 33% had prior chemotherapy, 62% had a disease-free interval of > 2 years, and 59% had only one metastatic site. The complete (14%) and partial (6%) response rate for 147 evaluable patients was 20% (95% CI for CR + PR of 14-27%). Multivariate analysis revealed that improved response was related to soft tissue involvement and positive PR status. The majority of patients with soft tissue, nodal or lung metastases had responses noted within three months. Median time to disease progression was 6.7 months. Multivariate analysis revealed that older patients, those with one metastatic site and those with positive PR status had the longest time to progression. Median survival was 27.2 months. Better performance status, fewer metastatic sites and being PR positive were associated with significantly improved survival. CONCLUSIONS The patient population in this series is not likely to be studied in future trials because of the wide use of tamoxifen in the adjuvant setting. In a small percentage of patients with metastatic breast cancer, tamoxifen therapy is associated with prolonged remission and survival. Pretreatment characteristics can help identify such patients.