Results from the standpoint of survival rates and locoregional failures are compared in three series of patients having had a radical mastectomy for breast cancer: (1) radical mastectomy alone for the patients who had essentially outer quadrant lesions and a negative axilla; (2) postoperative irradiation when the axillary nodes were positive and/or the tumor was centrally located or in the inner quadrants; and (3) preoperative irradiation for patients with an outside biopsy presenting with a very disturbed breast with edema and ecchymosis, and in a small group of patients with a lesion of clinically borderline operability. The ten-year survival rates are identical in the three groups. In the radical mastectomy alone group, 14% of the patients had positive axillary nodes, in the preoperative irradiation group 30% (probably one half of the true incidence without preoperative irradiation), and in the postoperative group, 71%. This data is indicative that irradiation, either pre- or postoperatively, has survival benefits since there is direct relationship between the percentage of patients with positive axillary nodes and the survival rates. However, there is no evidence that preoperative irradiation is superior to postoperative irradiation.