Clinical features, 8 histological variables, 7 nuclear morphometric variables, 2 mitotic indices, oestrogen-receptor (ER) and progesterone-receptor content (PR), DNA ploidy and S-phase fraction (SPF) were entered in a Cox's model to assess their independent predictive value in 216 breast-cancer patients followed up for over 9 years. In the whole series, histological type (p = 0.007), volume-corrected mitotic index (M/V index) (p = 0.01), axillary-lymph-node (pN) status (p = 0.024) and the year of treatment (p = 0.045) predicted independently the recurrence-free survival (RFS). In a sub-analysis including SPF (n = 148), the year of treatment (p = 0.003), tumour diameter (p = 0.004), SPF (p = 0.022) and nuclear pleomorphism (p = 0.056) independently predicted the RFS. In a Cox's analysis of the whole series, tumour diameter (p less than 0.001), pN status (p = 0.001), PR status (p = 0.002) and the year of treatment (p = 0.021) were independent predictors of survival. In a separate analysis including also SPF (n = 148), tumour diameter (p less than 0.001), SPF (p = 0.003), pN status (p = 0.008) and the year of treatment (p = 0.015) proved to be independent prognostic factors. The results show that tumour diameter, pN status, M/V-index, histological type, SPF and PR status comprise a sufficient combination of prognostic factors in female breast cancer. In pN patients, age and SDPE may be of additional prognostic significance. The prognostic scores combining the independent prognostic variables reflecting both the proliferative rate and metastatic potential of the tumours are accurate predictors of the RFS and overall survival.