Prior clinical studies have indicated that total nucleated and CD34+ cell dose is a critical determinant of hematopoietic reconstitution and survival after unrelated donor umbilical cord blood (UCB) transplantation. Efforts to overcome the problem of small cell numbers in UCB grafts by ex vivo expansion, primarily by culture with cytokines, have not met with success. We have previously shown that culture of CD34+CD38− UCB progenitors with the Notch ligand, Delta1, results in enhanced generation of NOD/SCID repopulating cells. Here we develop a clinically feasible cGMP method for Notch ligand-based expansion of cord blood precursors. Specifically, we investigated the use of CD34+ versus CD34+CD38− cells as a starting population, optimal cytokines and medium, selection of culture vessel and culture period for effects on generation of NOD/SCID repopulating cells. UCB progenitors were cultured in the presence of a Notch ligand form consisting of the extracellular domain of Delta1 fused to the Fc domain of human IgG1 (Delta 1ext-IgG) or control human IgG. Initial studies demonstrated optimal cytokines consisted of SCF, FL, TPO, IL6 and IL3, together with fibronectin fragments in serum free medium. There was no significant difference seen in the CD34 fold expansion with CD34+ versus CD34+CD38− starting cells, however, upon transplantation into NOD/SCID mice, there was a significant increase in the level of human engraftment seen with the CD34+ starting cell population (6.93% vs 2%; p=0.01). Further results from 5 independent experiments in which cord blood CD34+ progenitor cells were cultured for 17 days with immobilized Delta1extIgG or control resulted in a mean fold expansion of CD34+ cells of 230 (± 53) for the Delta cultured cells versus 65 (±31) for the control cultured cells (p=0.03). Delta cultured cells demonstrated significantly enhanced levels of human engraftment as measured by percent CD45 in the marrow of the animals at both 3 weeks (Delta1 15.5%, control 2.6%, uncultured 0.2%; p<0.0001) and at 9 weeks (Delta1 29.4%, control 8.9%, uncultured 7.3%; p<0.0001). We also found significantly greater numbers of SCID repopulating cells (SRC) detected 3 and 9 weeks following transplantation in the Delta1ext-IgG cultured cells compared to control cultured or non-cultured cells (frequency of SRC per 106: 3 weeks, 125 versus 37 or 8, respectively; p=0.0001; 9 weeks, 99 versus 56 or 15, respectively; p=0.01 and p=0.0001). Additional experiments demonstrated that the 17 day culture period was superior to shorter (10 days) or longer (21 days) periods. Relevant to anticipated administration of cultured cord blood units together with a second non-cultured unit in clinical trials, we determined the relative contribution to engraftment of co-infused expanded versus non-manipulated cells in immunodeficient mice, using tissue culture bags as a closed system. We found increased human engraftment in mice that received co-infusions of cultured and uncultured cells compared to either unit alone. Moreover, studies demonstrated that both units contributed to the observed human engraftment suggesting absence of cross-immunologic rejection. This data demonstrate the ability to ex vivo expand UCB repopulating cells using a clinically relevant Notch ligand-based closed culture system and suggests clinical evaluation of this approach to provide more rapid engraftment to overcome the major disadvantage of UCB transplantation.