Mammosphere and breast tumoursphere culture have gained popularity as in vitro assays for propagating and analysing normal and cancer stem cells. Whether the spheres derived from different sources or parent cultures themselves are indeed single entities enriched in stem/progenitor cells compared to other culture formats has not been fully determined. We surveyed sphere-forming capacity across 26 breast cell lines, immunophenotyped spheres from six luminal- and basal-like lines by immunohistochemistry and flow cytometry and compared clonogenicity between sphere, adherent and matrigel culture formats using in vitro functional assays. Analyses revealed morphological and molecular intra- and inter-sphere heterogeneity, consistent with adherent parental cell line phenotypes. Flow cytometry showed sphere culture does not universally enrich for markers previously associated with stem cell phenotypes, although we found some cell-line specific changes between sphere and adherent formats. Sphere-forming efficiency was significantly lower than adherent or matrigel clonogenicity and constant over serial passage. Surprisingly, self-renewal capacity of sphere-derived cells was similar/lower than other culture formats. We observed significant correlation between long-term-proliferating-cell symmetric division rates in sphere and adherent cultures, suggesting functional overlap between the compartments sustaining them. Experiments with normal primary human mammary epithelia, including sorted luminal (MUC1(+)) and basal/myoepithelial (CD10(+)) cells revealed distinct luminal-like, basal-like and mesenchymal entities amongst primary mammospheres. Morphological and colony-forming-cell assay data suggested mammosphere culture may enrich for a luminal progenitor phenotype, or induce reversion/relaxation of the basal/mesenchymal in vitro selection occurring with adherent culture. Overall, cell line tumourspheres and primary mammospheres are not homogenous entities enriched for stem cells, suggesting a more cautious approach to interpreting data from these assays and careful consideration of its limitations. Sphere culture may represent an alternative 3-dimensional culture system which rather than universally 'enriching' for stem cells, has utility as one of a suite of functional assays that provide a read-out of progenitor activity.