Two hundred and twenty-three patients with disseminated breast cancer entered in a randomized trial: Group I: 130 patients were given a monthly 5 day- course of a cytotoxic chemotherapy protocol including Adriamycine, Vincristine, Cyclophosphamide, 5 Fluoro-uracile (A.V.L.F.); Group II: 93 patients were given alternatively the same program and a non cross resistant program including: VM 26, Mitomycine C, Methotrexate (V.M.M.). The patients of these groups were randomized into three subgroups: Subgroup O: chemotherapy alone; Subgroup N: chemotherapy + Tamoxifen (20 mg/m2/day); Subgroup NN: chemotherapy + Tamoxifen + Norethisterone (40 mg/m2/day). Objective response rates to cytotoxic chemotherapy were respectively of 65 per cent and 68 per cent in groups I and II of chemotherapy. Further more mean-duration of response and survival were not different. Objective response rates were comparable in the three subgroups of chemo-hormonotherapy. However, the survival was significantly better in the subgroup N (with Tamoxifen) when compared with the subgroup O (without hormonotherapy), and marginally better (p = 0,09) when compared with the NN subgroup (with Tamoxifen + Norethisterone).