While cultured embryonic stem (ES) cells can be harvested in abundance and appear to be the most versatile of cells for regenerative medicine, adult stem cells also hold promise, but the identity and subsequent isolation of these comparatively rare cells remains problematic in most tissues, perhaps with the notable exception of the bone marrow. The ability to continuously self-renew and produce the differentiated progeny of the tissue of their location are their defining properties. Identifying surface molecules (markers) that would aid in stem cell isolation is a major goal. Considerable overlap exists between different putative organspecific stem cells in their repertoire of gene expression, often related to self-renewal, cell survival and cell adhesion. More robust tests of 'stemness' are now being employed, using lineage-specific genetic marking and tracking to show production of long-lived clones and multipotentiality in vivo. Moreover, the characterization of normal stem cells in specific tissues may provide a dividend for the treatment of cancer. The successful treatment of neoplastic disease may well require the specific targeting of neoplastic stem cells, cells that may well have many of the characteristics of their normal counterparts.