Autologous peripheral blood stem cell transplantation (PBSCT) is replacing autologous bone marrow transplantation (BMT) in the treatment of leukemia. One of the potential advantages of autologous PBSCT is the possibility that peripheral blood stem cells (PBSC) are less likely to be contaminated by leukemic cells than bone marrow grafts. However, the major problem still remains the high incidence of leukemic relapse following autologous PBSCT, which may be caused by the reinfusion of PBSC contaminated by leukemic cells. Recently, we have developed a quantitative assay using competitive reverse transcriptase polymerase chain reaction that estimates the number of AML1/ETO transcripts in t(8;21) acute myelogenous leukemia (AML), in order to determine the degree of leukemic cell contamination in PBSC harvests, and to monitor minimal residual disease (MRD) quantitatively in patients with t(8;21) AML. Our data indicate that although PBSC harvests collected after consolidation chemotherapy are contaminated by leukemic cells, the degree of leukemic cell contamination decreases with repeated cycles of chemotherapy. Furthermore, the MRD in PBSC harvests is less than in the corresponding bone marrow obtained on the day of the PBSC collection. There appears to be no relationship between the number of AML1/ETO transcripts found in the infused PBSC harvests and the incidence of leukemic relapse following autologous PBSCT in our study. However, a substantial decrease of AML1/ETO transcripts was seen following autologous PBSCT. Thus, the quantitative analysis of AML1/ETO transcripts may be clinically useful in patients with t(8;21) AML.