This study concerns 645 patients first treated with surgery for unilateral, non metastatic, invading breast cancer. Intratumoral estrogen receptor and progesterone level were determined in every case. Level greater than 5 fmoles/mg cytosolic protein was considered as positive for both receptors. Univariate analysis has pointed out a significant linkage between overall survival and the following factors: age, clinical size of the tumor, histopathological grading SBR, clinical and histological lymph node involvement, capsular tear, RO and RP status. Statistical significance of menopausal status is borderline. Number of tumor foci is not significant. Likewise, disease free survival was correlated to the same factors. Multivariate analysis (Cox), secondarily pointed out that overall survival was strongly related to age, size of the tumor, lymph node involvement, capsular tear, histopathological grading SBR, menopausal status and RP. Multivariate analysis of the disease free survival revealed that it was strongly related to age, tumor size, lymph node involvement, capsular tear, grading SBR and RP. An interaction has been pointed out between Ro and menopausal status: Ro greater than 5 fmoles/mg cytosolic protein carry its own prognostic weight (Cox) and lengthen overall survival only for post menopausal women. A prognostic score, taking into account all of these factors has been calculated for both overall survival and disease free survival, and enabled us to isolate 4 groups of patients with good, intermediate and bad prognosis. These 2 models have been validated on an independent group of patients according to the sample test procedure. This results indicate that hormonal receptors carry their own prognostic weight in operable breast cancer (only for postmenopausal women for RO), and should be taken into account when adjuvant therapies are indicated after surgical treatment for breast cancer, in association with other more usual prognostic factors.