CD34+ cells in human cord blood and marrow are known to give rise to dendritic cells (DC), as well as to other myeloid lineages. CD34+ cells are rare in adult blood, however, making it difficult to use CD34+ cells to ascertain if DC progenitors are present in the circulation and if blood can be a starting point to obtain large numbers of these immunostimulatory antigen-presenting cells for clinical studies. A systematic search for DC progenitors was therefore carried out in several contexts. In each case, we looked initially for the distinctive proliferating aggregates that were described previously in mice. In cord blood, it was only necessary to deplete erythroid progenitors, and add granulocyte/macrophage colony-stimulating factor (GM-CSF) together with tumor necrosis factor (TNF), to observe many aggregates and the production of typical DC progeny. In adult blood from patients receiving CSFs after chemotherapy for malignancy, GM-CSF and TNF likewise generated characteristic DCs from HLA-DR negative precursors. However, in adult blood from healthy donors, the above approaches only generated small DC aggregates which then seemed to become monocytes. When interleukin 4 was used to suppress monocyte development (Jansen, J. H., G.-J. H. M. Wientjens, W. E. Fibbe, R. Willemze, and H. C. Kluin-Nelemans. 1989. J. Exp. Med. 170:577.), the addition of GM-CSF led to the formation of large proliferating DC aggregates and within 5-7 d, many nonproliferating progeny, about 3-8 million cells per 40 ml of blood. The progeny had a characteristic morphology and surface composition (e.g., abundant HLA-DR and accessory molecules for cell-mediated immunity) and were potent stimulators of quiescent T cells. Therefore, large numbers of DCs can be mobilized by specific cytokines from progenitors in the blood stream. These relatively large numbers of DC progeny should facilitate future studies of their Fc epsilon RI and CD4 receptors, and their use in stimulating T cell-mediated resistance to viruses and tumors.