One hundred fifty-four patients being treated for recurrent breast cancer were studied for response rates and prognosis according to first recurrent lesions in radiation, chemotherapy, hormone therapy and combined therapy groups. It was important to consider the sites of the recurrent lesions in the treatments. Although there was no significant difference in prognosis among the first recurrent therapy groups, the overall cumulative five-year survival rate of the patients who responded (to the initial treatment) was significantly better than that of those who did not. As for the treatment of recurrent breast cancer, it is important to consider hormone receptor, lesion and age and to choose the effective mode of treatment for the first (recurrence of the) disease.