Hormone-replacement therapy (HRT) is widely used by post-menopausal women. Although this treatment may slightly increase the incidence of breast cancer, more and more cases are diagnosed while women are taking HRT. The purpose of this study was to ascertain the influence of HRT on prognostic factors and outcome of breast cancer. Data on all breast-cancer patients, including precise information on HRT, was prospectively and systematically recorded in a data base. From 1985 to 1995, 1379 post-menopausal women fulfilled the eligibility criteria for this study. All were treated by us (P.B. and L.P.) in our ward of a large public hospital of Marseilles, France. The clinical features, laboratory findings and survival rates in 142 HRT users who developed breast cancer while being treated were compared with those of 284 matched never user breast-cancer patients. Patients who developed breast cancer during HRT had fewer locally advanced cancers and smaller and better-differentiated cancers. Lymph-node involvement was significantly less frequent in the user group than in the non-user group (non-significant). Estradiol receptivity was both qualitatively and quantitatively lower in users. There was no significant difference with regard to recurrence and metastasis-free survival and overall survival. We conclude that HRT does not affect the prognosis of breast cancer. Regular surveillance during HRT allows early detection of smaller lesions. The higher number of well-differentiated cancers and the distribution of hormone receptivity may reflect interaction between neoplastic tissue and exogenous hormones.