The size of a breast cancer at diagnosis has conventionally been thought of as a fundamental and critical determinant of clinical outcome. However, the tendency of some subtypes of breast cancer to behave aggressively, despite being small (</=1 cm in diameter), questions the premise that cancer size should always be considered in treatment decisions. Although there is an association between tumor size and lymph-node involvement for most tumor types, this pattern is not invariable. We speculate that the uncoupling of tumor size, lymph-node status and prognosis in some subtypes of breast cancers might reflect an underlying disproportionate relationship between the number of cancer cells with metastatic potential and the size of the cancer. Alternatively, some small cancers might harbor cells that are inherently aggressive and are likely to metastasize. These observations have implications for the screening and treatment of breast cancers, particularly for women with basal-like and BRCA1-related breast cancers.