The precise cell types that give rise to tumors and mechanisms that underpin tumor heterogeneity are poorly understood. There is increasing evidence to suggest that diverse solid tumors are hierarchically organized and may be sustained by a distinct subpopulation of cancer stem cells (CSCs). The CSC hypothesis provides an attractive cellular mechanism that can account for the therapeutic refractoriness and dormant behavior exhibited by many tumor types. Breast cancer was the first solid malignancy from which CSCs were identified and isolated. Direct evidence for the CSC hypothesis has also recently emerged from mouse models of mammary tumorigenesis, although alternative models to explain heterogeneity also seem to apply. Our group has found that the luminal epithelial progenitor marker CD61/beta3 integrin identified a CSC population in mammary tumors from MMTV-wnt-1 mice. However, no CSCs could be identified in the more homogeneous MMTV-neu/erbB2 model, suggesting an alternate (clonal evolution or stochastic) model of tumorigenesis. It seems likely that both paradigms of tumor propagation exist in human cancer. From a clinical perspective, the CSC concept has significant implications. Quiescent CSCs are thought to be more resistant to chemotherapy and targeted therapy. Enrichment of putative CSCs has been noted in studies of chemotherapy-treated patients, lending support to the CSC hypothesis and their potential role in chemoresistance. Although many unresolved questions on CSCs remain, ongoing efforts to identify and characterize CSCs continue to be an important area of investigation, with the potential to identify novel tumor targeting strategies.