Primitive haematopoietic progenitor and stem cells (HSC) have been pursued as highly desirable targets for genetic therapy as technology allowing safe and controllable transfer of exogenous genes into eukaryotic cells was developed a decade ago. Retroviral vectors have been used for the majority of preclinical and clinical studies directed at these cells, because these vectors have a number of the necessary properties, including chromosomal integration, helper-free production systems, and lack of toxicity. Until recently, however, results with these vectors in clinical trials and large animal models indicated efficiency of gene transfer as a major hurdle to be overcome. We have focused on using the rhesus macaque autologous transplantation model to optimize gene transfer to primitive haematopoietic cells, and investigate questions regarding in vivo stem cell behaviour, in a system with proven predictive value for human haematopoiesis. By optimization of transduction conditions using standard vectors, gene transfer efficiency to primitive repopulating cells has reached the clinically relevant range of 5-20% long-term. Alternative vector systems, have also yielded promising results. We have also found that relatively simple manipulation of cell cycle status prior to reinfusion of marked cells results in significantly improved engraftment of transduced cells: this finding may have an impact particularly in the nonablative setting. The high level marking has permitted insertion site analysis and clonal tracking in vivo. Inverse PCR and/or a ligation-mediated PCR procedure have demonstrated that a large number of transduced clones (over 50) contribute to multiple lineages in vivo for up to at least 2 years post-transplantation. Thus far we have little evidence for rapid clonal succession or lineage-restricted engraftment of transduced cells. These and other advances should result in successful gene therapy for a variety of acquired and congenital disorders affecting HSCs and their progeny lineages.