Recent studies in the NOD/SCID model have shown improved engraftment of SCID-repopulating cells and higher levels of engraftment in the secondary transplantation when cells were administered by intramarrow (IM) versus intravenous (IV) injection suggesting that direct injection into the marrow cavity may be beneficial for stem cell engraftment in a clinical setting. To study whether IM injection was feasible and would result in improved engraftment in a clinically relevant large animal model, we compared IM vs IV injection in our competitive repopulation assay in baboons. Enriched CD34+ cells were split into 2 equal fractions and transduced with either a GFP- or YFP-expressing vector. Pretransplant transduction efficiencies and expansion of CD 34+ cells were similar in both fractions. One fraction was then infused into the marrow cavity of the right femur and the other fraction was given intravenously. Three baboons received gene-modified CD34+ enriched autologous bone marrow cells after myeloablative radiation. Peripheral blood granulocyte marking levels showed peaks at 2–3 weeks after transplantation and decreased thereafter. In all three monkeys, marking levels of IM injected cells (GFP) were lower than marking levels of IV injected cells (YFP) early after transplantation up to 7 weeks. However, in two of the three monkeys, GFP marking increased steadily after 2 months resulting in higher marking levels from IM injected cells. The trend sustained up to the last follow-up of nine months after transplantation, marking levels being 25.5% and 7.4% from IM and IV injected cells, respectively, in M00228. This pattern was recapitulated in the marking of bone marrow cells of the two animals. GFP (IM) and YFP (IV) marking levels of bone marrow cells from non-injected bone were 24.2% and 33.9%, respectively, at 1 month, 7.9% and 4.6% at 3 months, 19.1% and 12.6% at 6 months after transplantation in M00228. In addition, the GFP marking of the bone marrow cells from the injected bone was higher than that of the BM cells from non-injected bone while YFP marking level was similar. In conclusion, our data suggest that direct intramarrow injection of CD34+ cells may lead to improved engraftment of long-term repopulating cells. Clonal analysis is currently under way to determine the clonal pattern of the differentially marked repopulating cells.