Recently, RD114 (feline endogenous retrovirus envelope protein)-pseudotyped retroviral particles have been shown to transduce human NOD/SCID repopulating cells efficiently. In this study, we compared directly transduction of repopulating cells with RD114-pseudotyped vector to that with standard amphotropic vector in the rhesus macaque model. G-CSF/SCF-mobilized CD34(+) rhesus peripheral blood cells were cultured in the presence of SCF, Flt-3 ligand, and MGDF on Retronectin-coated flasks. To assess directly the ability of the two pseudotypes to transduce primitive cells, both vectors were added simultaneously to the target cells every 24 h, for a total of four exposures in 96 h. The cells were reinfused after the animals received 1000 cGy total body irradiation. At the end of transduction, gene marking efficiency of CFU was higher with amphotropic LNL6 vector (mean 88.4%) vs RD114-G1Na vector (mean 18.5%). After long-term engraftment in three animals, total neo gene marking levels were 4-5% in PBMNCs and 1.5-4% in granulocytes. The RD114-G1Na marking levels were consistently higher in granulocytes than in mononuclear cells, while amphotropic LNL6 marking levels were higher in PBMNCs than in granulocytes. The differential gene marking patterns suggest that RD114 and amphotropic vectors may target distinct progenitor or stem cell populations. There was no clear advantage for RD114-pseudotyped vectors in this predictive preclinical model in terms of overall long-term marking levels; however, optimization of transduction conditions by increasing m.o.i. or inducing the receptor could potentially improve results with this novel vector system.