From 1976 to 1984, 427 postmenopausal women with high-risk breast cancer (pN + or pT greater than 30 mm) were randomized between postoperative radiation therapy (RT), radiation therapy plus tamoxifen (RT-TAM), adjuvant chemotherapy (CT), or chemotherapy plus tamoxifen (CT-TAM). Surgery was a modified radical mastectomy in all cases. The radiation therapy was given with high-voltage techniques and included the chest wall and regional nodes. The dose was 4600 cGy/4 1/2 weeks. Tamoxifen was given at a dose of 40 mg daily for 2 years. The adjuvant chemotherapy consisted of 12 cycles of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) (or chlorambucil, methotrexate, and 5-fluorouracil [LMF] for patients entered before 1978). At a median follow-up time of 6 1/2 years the recurrence-free survival was significantly better for patients allocated to radiation therapy compared to chemotherapy and for patients allocated to tamoxifen compared to no adjuvant endocrine treatment (P less than 0.01). At 10 years the recurrence-free survival for patients in the RT-TAM, RT, CT-TAM, and CT groups was 63%, 57%, 47%, and 31%, respectively. A significant reduction of treatment failures with tamoxifen was only observed among patients with estrogen receptor-positive tumors. The overall survival difference in favor of patients allocated to radiation therapy or tamoxifen was not significant: the respective survival percentage at 10 years in the RT-TAM, RT, CT-TAM, and CT group was 65%, 62%, 52%, and 50%. The results indicate that postoperative radiation therapy continues to play an important role in the primary management of postmenopausal women with high-risk breast cancer and that the addition of tamoxifen may further improve the results among ER-positive patients.