Introduction of genes conferring drug resistance into hematopoietic cells may allow for improved chemotherapy by protection of normally drug-sensitive cells from the toxic side-effects of antitumor agents. We recently reported that transplantation of murine marrow transgenic for drug-resistant dihydrofolate reductase (DHFR) protected mice from lethal doses of methotrexate (MTX), demonstrating the feasibility of imparting drug resistance to recipients of marrow expressing drug-resistant DHFR activity. In order to optimize this strategy it is necessary to identify the hematopoietic cell populations which mediate drug resistance. For this purpose, we separated committed progenitor populations from primitive hematopoietic cells in DHFR transgenic marrow by counterflow elutriation (CE). As expected, supplementation with a fraction containing committed progenitors afforded protection from MTX-induced aplasia observed early after transplantion in animals administered MTX. In contrast, supplementation with a fraction containing primitive hematopoietic cells depleted of committed progenitors failed to provide immediate protection from early aplasia, but instead contributed to drug resistance 4 to 5 weeks after transplantation. The presence of primitive hematopoietic progenitors in both fractions was evident from Southern hybridization analysis for donor transgenic cells 2 months post-transplant. We conclude that protection from aplasia associated with MTX administration immediately after transplantation requires expression of drug-resistant DHFR activity in more differentiated, committed hematopoietic cells, while primitive DHFR transgenic progenitors contribute to long-term drug resistance. These results help to define the appropriate target cells for improved chemotherapy by protection of hematopoietic cells through the introduction of new genes conferring drug resistance.