Human bone-marrow or cord-blood progenitors (i.e., CD34+ cells) are easily purified by immunological methods and can be cultured on normal human-bone-marrow stromal cells for limited periods of time. Under these culture conditions, the number of progenitors declines in a few weeks and these cells disappear completely in less than 8 weeks. This fact suggests that this culture system is deprived of growth factor(s) able to support the self-renewal of stem cells. We have developed the culture of immunomagnetically purified human-bone-marrow- or cord-blood-derived CD34+ cells on a supportive mouse lipoblastic stromal cell line, MS-5. The long-term survival of clonogenic cells was analyzed in these cultures and compared with the results obtained by culture on human-bone-marrow stromal cells. The results demonstrated that only coculture of CD34+ cells on MS-5 layers allows the survival of clonogenic progenitors for at least 12 weeks. Cytospin smears were regularly performed and cell morphology was examined after classical staining methods (i.e., M.G.G. and toluidine blue staining). Histologic analysis demonstrated the growth of mast-cell-like metachromatic cells after the second week of incubation on MS-5 layer. The highest percentage of these cells was observed after 8 weeks, and averaged about 30 percent for cord-blood cells and 70 percent for bone-marrow cells. To further confirm the nature of the metachromatic cells obtained under this culture condition, immunohistochemical staining of tryptase was performed on the same samples. The results demonstrated similar percentages of tryptase+ cells and of metachromatic elements. Measurement of cellular histamine demonstrated that culture of CD34+ cells on MS-5 monolayers induced the formation and increase of this mediator. To determine whether the contact between MS-5 layers and CD34+ cells was an absolute requirement for the development of mast cells, CD34+ cells were cultured in the presence of MS-5 conditioned medium. This condition allowed the development of similar percentage of mast cells when compared with the coculture experiments, indicating that a soluble factor was involved in mast cell differentiation. Whatever the soluble factor(s) responsible for this mast cell growth activity, our culture system allows us to obtain significant amounts of highly enriched normal human mast cell populations useful for further studies on the reactivity of this cell subset.