Transfer of the multidrug resistance-1 (MDR1) gene to hemopoietic cells for myeloprotection against cytostatic agents is a new and rapidly developing field in "cancer gene therapy." Before clinical application, safety and efficacy criteria need to be met. The retroviral producer cell lines and the retroviral supernatant need to be tested for replication-competent retrovirus and contamination with adventitious agents. The cell source needs to contain sufficient hemopoietic cells with repopulating ability. We used CD34(+)-selected mobilized peripheral blood progenitor cells (PBPC) for MDR1 transductions in order to obtain a favorable vector to target cell ratio. An analysis of 249 patients who had undergone PBPC harvesting revealed that primarily solid tumor and non-Hodgkin's lymphoma patients are eligible for CD34+ selection. They can be expected to retain sufficient CD34+ cells for rapid and sustained engraftment after myeloablative therapy if the CD34+ cell loss (approximately 50%) during the procedure is taken into account. Clinical MDR1 gene therapy protocols focus on these two patient groups. Next we characterized MDR1 gene transfer into lineage-committed and primitive hemopoietic cells. Provirus-specific polymerase chain reactions showed a high efficiency gene transfer into colony-forming-units granulocyte-macrophage and long-term culture cells. The level of the conferred P-glycoprotein expression was estimated by fluorescence-activated cell sorting analysis to be up to 3 log above mock-transduced controls. The cobblestone area forming cell assay, which is a stroma-dependent long-term culture assay measuring frequencies of stem cell subsets in a limiting-dilution set-up, allowed demonstration of sustained expression of the MDR1 gene in the progeny of primitive hemopoietic cells. This is a favorable basis for a clinical MDR1 gene therapy trial.