Patients selected for neo-adjuvant therapy in this trial suffered from breast cancer stages T3/4, N0/+, M0. Selected patients were stratified and assigned to either pre- and post-operative chemotherapy (tamoxifen, fluorouracil, vincristine, methotrexate, cyclophosphamide, mitoxantrone) or radiotherapy by randomization. This interim report deals with side-effects and response rates of pre-operative therapy only. For this subgroup from 41 centres, 67 patients have so far been accepted for protocol treatment. 36 received peri-operative chemotherapy and 31 perioperative radiotherapy. We have completed reports of 66 pre-operative chemotherapeutic cycles pointing out only low side-effects in spite of a combination consisting of 5 different cytotoxic agents plus hormonal therapy. It is remarkable that 79% had no hair loss and 74% no gastrointestinal side-effects; in the chemotherapy-group 24 patients so far have been operated. Response rates were as follows: 1 (4%) complete remission, 9 (38%) partial remission, 7 (29%) minor response, 2 (8%) no response and 2 (8%) progression. In 2 patients response could not be determined. In the perioperative radiotherapy-group 10 patients have been operated with similar response rates. It is interesting to note, that both therapies influenced receptor values in different ways: Chemotherapy decreased receptor values, whereas radiotherapy insignificantly increased them. The purpose of this study is to establish a rational base for post-operative cytotoxic treatment of patients as only patients who objectively respond to preoperative treatment are likely to benefit from postoperative drug administration. The study is open for enrollment until a total of 200 patients is reached.