Villadsen R. In search of a stem cell hierarchy in the human breast and its relevance to breast cancer evolution. APMIS 2005;113:903–21. By deliberate analogy with the well-established concept of hematopoiesis, the term “mammopoiesis” is occasionally used to describe the development of the different cellular lineages and functional units in the mammary gland. The use of this term signifies a strong bias towards the idea that tissue homeostasis during mammary development, pregnancy, lactation and involution is brought about by the action of somatic stem cells characterized by longevity and multipotency. The progenies hereof eventually differentiate into structurally and functionally well-defined ductal-lobular units. During the past two decades evidence of such a notion in the mouse has developed from being largely circumstantial based on non-clonal in vivo experiments to a quite elaborate characterization of individual candidate stem cells by a number of different properties. Within tumor biology this has led to a renaissance of the concept of tumors as caricatures of tissue renewal. Thus, recent molecular classification of breast cancer based on genome wide expression analysis operates with different subtypes with specific reference to the normal luminal epithelial and myoepithelial/basal lineages in the breast. Apparently some tumors are lineage restricted and others differentiate more broadly as if they have preserved some stem-like properties. This holds promise for the existence of a stem cell hierarchy, the understanding of which may prove to be instrumental in further dissecting the histogenesis of breast cancer evolution. Most attention has been devoted to the question of different cellular origins of cancer subtypes and different susceptibilities of possible stem cells to gain or loss of oncogenes and tumor suppressor genes, respectively. Invaluable progress has been made over the past two decades in culture technology not only in terms of population doubling and clonal growth, but also the availability of lineage specific markers, cell sorting, and three-dimensional functional assays for tissue specific morphogenesis. Transcriptional profiling of stem cell zones has unraveled a hitherto unknown preservation of signaling pathways for maintenance of stem cell properties across tissue boundaries and species. Somatic stem cells have therefore been narrowed down to specific anatomic locations not only in rapidly renewing tissues such as skin and skin derivatives, but also in tissues with slower turnover times, such as lung, kidney and prostate. It is therefore now possible to integrate this information in a search for similar cells within the breast. Even if cell turnover after birth is provided exclusively by dividing lineage-restricted cells, more information about the robustness of breast differentiation programs during tumor progression is still very much required. Complete knowledge of the primary cell of origin of breast cancer and the mechanisms that influence differentiation programs during tumor initiation, promotion and progression may be crucial for the development of novel non-toxic therapies that influence tumor cell behaviour. The scope of this review is to discuss reports that have begun to elucidate the topographic location, key cellular type and lineage fidelity in culture and xenograft models of candidate human breast stem cells and their differentiated progenies with particular emphasis on comparison with the differentiation programs of tumor subtypes.