Nine hundred and eighty-four breast cancer patients were interviewed regarding exogenous hormonal use. This represents a random sample of breast cancer patients in Southern Sweden referred to the Department of Oncology at Lund for treatment between 1978 and 1997 (excluding 1980 and 1981) with a 100% follow-up. Ever-use of hormone replacement therapy (HRT) prior to diagnosis was significantly associated with a longer overall survival in women with their breast cancer diagnosed at ages 45 and above, relative risk (RR) of dying 0.73 (95% confidence interval (CI) 0.62-0.87; P = 0.0005). Ever use of HRT prior to breast cancer diagnosis was significantly positively associated with overall longer survival after adjustment for T-stage, N-stage, M-stage, year of diagnosis and age at diagnosis, RR of dying 0.78 (95% CI 0.65-0.93; P = 0.006). Hormone replacement therapy use and oestrogen receptor positivity were independently significantly associated with overall longer survival, P = 0.005 and P < 0.0001, respectively, in one model. HRT use and progesterone receptor positivity were also independently significantly associated with longer overall survival, P = 0.003 and P = 0.0003, respectively, in another model. The mode of diagnosis was known in 705 women. Mammography screening was not more common among HRT users compared with never-users, where this information was available. Both mammography screening and HRT use were independently associated with longer survival, P = 0.002 and P = 0.038 respectively.