Two hundred patients with node positive stage II breast cancer were randomised to four groups after radical mastectomy and axillary evacuation: (1) Postoperative radiotherapy, (2) Adjuvant chemotherapy with eight courses of CAFt (cyclophosphamide 500 mg m-2 + doxorubicin 40 mg/m-2 + ftorafur 20 mg kg-1 orally day 1-14) every fourth week, (3) Postoperative radiotherapy and adjuvant chemotherapy and (4) postoperative radiation, adjuvant chemotherapy and tamoxifen 40 mg daily for 2 years. Thirty-two per cent of the patients discontinued treatment due to GI-toxicity, while 26% required dose reductions due to leukopenia. Radiation pneumonitis was more frequent after the combination of postoperative radiotherapy with chemotherapy. There was a better relapse-free survival in the groups receiving chemotherapy compared to radiotherapy alone (P = 0.05), which was highly significant in a multivariate Cox analysis (P = 0.004). No significant survival differences were seen. Tamoxifen had no clear overall effect but there were better relapse-free (P = 0.04) and overall (P = 0.004) survival with tamoxifen in estrogen receptor positive patients, while estrogen receptor negative patients had a somewhat poorer survival (P = 0.07) after tamoxifen. Local control was better (NS) after the combination (93%) radiotherapy and chemotherapy compared to either treatment alone (76% with radiotherapy and 74% with chemotherapy at 5 years).