Assignment of breast cancer patients to specific prognostic groups has been a relatively empty exercise until the recent acceptance of the effectiveness of systemic chemotherapy. The previous question of who should be so treated has changed in North America to a question of who should not be so treated. The availability of intensive chemotherapy for the worst prognostic groups as well as efficacious, low-toxicity adjuvant chemotherapy has made prognostication mandatory. The current major question is which women with breast cancer will not benefit from chemotherapy. We are actually much better at prognostication than many studies and the current broad search for new prognostic markers would indicate. The assignment of excellent prognosis categories by defined special histologic types or grade of breast cancer as well as other measures recognize women with survival approaching or equaling the general population. Likewise, extremely poor prognosis may be recognized in 20% to 40% of women with breast cancer by the presence of extensive mitoses or high proliferation index (by any modality) and poor differentiation (highest grade). Combined tumor size and nodal status data with high-grade indicates that women with these indicators will die within 2 to 3 years with as high as 80% certainty. The surgical pathologist still determines if cancer is present and must also determine if the tumor is in situ or invasive, its type or grade if invasive, and the extent of in situ component. Considering the diminution in overall size of cancers detected by modern techniques, it is clear that validation of newer prognostic variables will be more useful and will allow more separate determinations if adapted to tissue sections rather than tissue homogenates. The many measures of these carcinomas, such as size and nodal status, make multiparametric assessment of putative prognostic markers mandatory.