To evaluate the efficacy of radiotherapy without surgery, treatment results in patients treated for locally advanced breast cancer (n = 209) and those selected by positive axillary apex biopsy (n = 289) in the period between 1977 and 1985 have been analysed retrospectively. Treatment consisted of primary irradiation to the breast and regional lymph nodes followed by a boost to the primary breast tumour and palpable regional disease to a mean normalised total dose (NTD) of 64.7 Gy with a range of 33.4-93 Gy (2 Gy fractions, alpha/beta = 5 Gy). Adjuvant systemic treatment was given in 30% of the locally advanced and in 40% of the apex positive patients. Thirty percent of the apex positive patients had an excisional biopsy of the breast tumour. By multivariate analysis a prognostic index is constructed for locoregional control, overall survival and distant disease-free interval. Primary tumour size and clinical nodal status are independent prognostic factors for locoregional control. Based on the prognostic index for local control four different groups can be identified with 5 year local control rates varying from 47 to 86%. Patients treated with adjuvant chemotherapy and patients irradiated to a NTD of 60 Gy or more had significantly better local control. For overall survival primary tumour size, clinical nodal size and age are independent prognostic factors. Patients irradiated to a NTD above 60 Gy had significantly better results. Survival according to the prognostic index for survival varies between 20 and 50% at 5 years for the four groups subdivided according to the index for survival. Primary tumour size, clinical node size and age are independent prognostic factors for distant disease-free interval. Patients treated with adjuvant hormonal therapy had significantly better results. In the four groups subdivided according to the prognostic index for distant disease-free interval results vary from 17 to 30% at 5 years.