Ductal carcinoma in situ (DCIS) now accounts for 20-30% of all newly diagnosed breast cancers in centers which use mammographic surveillance as a standard part of the examination. The majority of these DCIS lesions, at least in the United States, are of very limited size, with mean estimated extents of 8-20 mm based on pathological examination. A small fraction of these are incidental microscopic features of the biopsy; the majority are detected on the basis of mammographic microcalcifications. These mammographically detected DCIS lesions are biologically heterogeneous, and this is reflected by their histology. Moreover, a number of recent independent studies have shown that the clinical outcome of patients, particularly those treated by breast conservation, is related to the presence of reproducible and identifiable histologic features, and possibly to certain immunohistochemically demonstrable gene markers as well. Nuclear and proliferative features of DCIS, as well as an evaluation of the periductular stroma, will be major focuses for the surrogate endpoint biomarker proposal.