Adoptive transfer of genetically modified somatic cells will play an increasingly important role in the management of a wide spectrum of human diseases. Among the most appealing somatic cells as potential gene transfer vehicles are hematopoietic cells, because of their wide distribution and their well-characterized capacities for proliferation, differentiation, and self-renewal. Genes can be readily transferred into short-lived and lineage-restricted hematopoietic cells, but there remains a need to develop reliable methods for gene transfer into hematopoietic stem cells in large animals. In this work, we used a gene transfer approach in which hematopoietic cells in long-term marrow cultures were exposed to the replication-defective retrovirus N2, bearing the reporter gene neo, on multiple occasions during 21 days of culture. Genetically marked cultured autologous cells were infused into 18 canine recipients in the absence of marrow-ablative conditioning. neo was detected by Southern blotting and/or the polymerase chain reaction in the marrow, blood, marrow-derived granulocyte/macrophage and erythroid progenitors, and cultured T cells in dogs after infusion. In most dogs, the proportion of long-term marrow culture cells contributing to hematopoiesis rose during the first 3 months after infusion and peaked within the first 6. The maximal levels attained were between 10% and 30% G418-resistant (neo-positive) granulocyte/macrophage progenitors. At 12 months, five dogs maintained greater than 10% G418-resistant progenitors, and for two of them this level exceeded 20%. Two dogs had greater than 5% G418-resistant hematopoietic progenitors at 24 months after infusion. Our data suggest that very primitive hematopoietic progenitors are maintained in long-term marrow cultures, where they can be triggered into entering the cell cycle. In vivo, these activated cells likely continue normal programs of proliferation, differentiation, and self-renewal. Their progeny can be maintained at clinically relevant levels for up to 2 years without the requirement that endogenous hematopoiesis be suppressed through chemo- or radiotherapy prior to adoptive transfer. Long-term marrow culture cells may thus be ideal targets for gene therapy involving adoptive transfer of transduced hematopoietic cells.