Many experimental and clinical protocols are being developed that involve ex vivo culture of human hematopoietic cells on stroma or in the presence of cytokines. However, the effect of these manipulations on primitive hematopoietic cells is not known. Our severe combined immune-deficient mouse (SCID)-repopulating cell (SRC) assay detects primitive human hematopoietic cells based on their ability to repopulate the bone marrow (BM) of immune-deficient non-obese diabetic/SCID (NOD/SCID) mice. We have examined here the maintenance of SRC, colony-forming cells (CFC), and long-term culture-initiating cells (LTC-IC) during coculture of adult human BM or umbilical cord blood (CB) cells with allogeneic human stroma. Transplantation of cultured cells in equivalent doses as fresh cells resulted in lower levels of human cell engraftment after 1 and 2 weeks of culture for BM and CB, respectively. Similar results were obtained using CD34⁺-enriched CB cells. By limiting dilution analysis, the frequency of SRC in BM declined sixfold after 1 week of culture. In contrast to the loss of SRC as measured by reduced repopulating capacity, the transplanted inocula of cultured cells frequently contained equal or higher numbers of CFC and LTC-IC compared with the inocula of fresh cells. The differential maintenance of CFC/LTC-IC and SRC suggests that SRC are biologically distinct from the majority of these in vitro progenitors. This report demonstrates the importance of the SRC assay in the development of ex vivo conditions that will allow maintenance of primitive human hematopoietic cells with repopulating capacity.