We have studied the role of different conditioning regimens for engraftment of genetically marked hematopoietic repopulating cells in dogs. Peripheral blood (PB) and/or marrow cells collected after treatment with recombinant canine stem cell factor (rcSCF) or cyclophosphamide were transduced in a vector-containing long-term culture system. Three different vector-producing cell lines with similar viral titers were used. In two of them, the neo-containing LN vector was packaged either in the PA317 cell line with an amphotropic murine retrovirus envelope or the PG13 cell line with the gibbon ape leukemia virus (GALV) envelope. The MFG/GC vector produced in PA317 cells contained the human glucocerebrosidase gene. Nineteen dogs received either no conditioning (group A, n = 5), irradiation to both humeri with 1,000 cGy (group B, n = 5), a sublethal dose of cyclophosphamide 40 mg/kg (group C, n = 4), a sublethal dose of 200 or 300 cGy total body irradiation (TBI) (group D, n = 3), or an otherwise lethal dose of 920 cGy TBI (group E, n = 3) before intravenous (groups A, C, D, E) or intramedullary (group B) infusion of the transduced autologous hematopoietic cells. Transduction efficiency of hematopoietic cells at the time of infusion into the animals was similar among the different conditioning groups. Dogs were observed for at least 6 months. PB granulocytes were obtained at least every 3 weeks after transplant and analyzed by polymerase chain reaction for the presence of the transduced genes. The percentages of positive results in dogs more than 4 weeks after transplantation were 0% without conditioning, 5% with local irradiation, 18% with sublethal cyclophosphamide, 33% with sublethal TBI, and 17% with otherwise lethal TBI. Analyzing the influence of conditioning regimens by a generalized estimating equation (GEE) technique, which considered the use of different retrovirus vectors and the number of mononuclear cells infused as potential confounding variables, we found that engraftment of genetically marked repopulating cells was significantly improved (P < .001) in dogs receiving systemic conditioning with either otherwise lethal TBI, sublethal TBI, or sublethal cyclophosphamide compared to dogs with local irradiation only or no conditioning. Within the limitation of the experimental design, these data suggest that myeloablative or myelosuppressive conditioning improves engraftment of genetically marked hematopoietic repopulating cells.