A statistical analysis was performed on a series of 170 consecutive cases of operable (M0) breast cancer in males. All the patients underwent surgery. The end-points considered were: (i) overall mortality, (ii) all neoplastic events and deaths without evidence of breast disease (first event). Five- and ten-year overall mortalities were 26.9 and 54.3%, respectively. A multiple regression analysis showed that tumour size and nodal status (pT and pN) were statistically significant as prognostic factors. With regard to first events, 12 local recurrences (thoracic wall), one nodal relapse in the axilla and one contralateral tumour were observed. Primary tumours, other than breast cancer, occurred in 11 patients. The observed probability of surviving at 10 years from the treatment was definitely lower than that of the general population. For the follow-up periods of 0-5 and 6-10 years, the excess death rate per 100 man-years was 9.98 and 13.43, respectively. It appears from the analysis that prognosis of breast cancer is worse in men than in women.