BACKGROUND: According to the overview of Early Breast Cancer Trialists' Collaborative Group, anthracycline containing regimens are superior to cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) as adjuvant chemotherapy for breast carcinoma, but no comparative information is available in terms of primary chemotherapy. In the current randomized controlled trial, the authors compared CMF with a chemotherapy regimen including CMF, epirubicin, and vincristine (CMFEV). METHODS: Two hundred eleven patients with Stages I and II palpable breast carcinoma and tumor diameter > 2.5 cm or < or = 2.5 cm with cytologically proven axillary lymph node involvement were randomized to receive CMF (arm A) or CMFEV regimen (arm B) for four cycles before surgery. After surgery, patients in both arms received adjuvant CMF for three cycles; the postmenopausal patients also received tamoxifen for two years. RESULTS: There were no significant differences in the complete response (CR) and in the CR plus partial response (PR) rates between the two arms. In the subset analysis, among premenopausal patients, significantly higher rates of CR (26% vs 4%, P = 0.004) and of CR + PR rates (80% vs 54%, P = 0.007) were observed in the CMFEV, as compared to the CMF arm. Multivariate analysis confirmed the presence of a significant interaction between menopausal status and type of treatment on the probability of achieving CR (P = 0.02) or CR + PR (P = 0.01). There were no major differences in the side effects of the two treatments, with the exception of more frequent alopecia in the experimental arm. CONCLUSIONS: The results of the current study are in line with those of previous published randomized clinical trials comparing regimens without and with anthracycline as adjuvant treatment, indicating an agreement between the short term response to primary chemotherapy and the long term results observed in the adjuvant setting.