Since only about half of primary breast cancer patients are cured by local treatment, factors which can predict which patients are likely to recur and should thus receive preventive treatmrnt are needed. Here, work on such prognostic factors which has been going on in San Antonio for many years will be reviewed. The significance of the number of involved axillary nodes and of steroid receptors is already well known, while the value of tumor proliferative rate and ploidy has been appreciated more recently, and the significance of amplification of the oncogene HER-2/neu is just now emerging.