Background Chimerism analysis is an important tool for the surveillance of post-transplant engraftment.It offers the possibility of identifying impending graft rejection and recurrence of underlying malignant or non-malignant disease.Here we investigated the quantitative chimerism kinetics of 21 relapsed leukemia patients after allogeneic hematopoietic stem cell transplantation (HSCT).Methods A panel of 29 selected sequence polymorphism (SP) markers was screened by real-time polymerase chain reaction (RT-PCR) to obtain the informative marker for every leukemia patient.Quantitative chimerism analysis of bone marrow (BM) samples of 21 relapsed patients and 20 patients in stable remission was performed longitudinally.The chimerisms of BM and peripheral blood (PB) samples of 14 patients at relapse were compared.Results Twenty-one patients experienced leukemia relapse at a median of 135 days (range,30-720 days) after transplantation.High recipient chimerism in BM was found in all patients at relapse,and increased recipient chimerism in BM samples was observed in 90％ (19/21) of patients before relapse.With 0.5％ recipient DNA as the cut-off,median time between the detection of increased recipient chimerism and relapse was 45 days (range,0-120 days),with 76％ of patients showing increased recipient chimerism at least 1 month prior to relapse.Median percentage of recipient DNA in 20 stable remission patients was 0.28％,0.04％,0.05％,0.05％,0.08％,and 0.05％ at 1,2,3,6,9,and 12 months,respectively,after transplantation.This was concordant with other specific fusion transcripts and fluorescent in situ hybridization examination.The recipient chimerisms in BM were significantly higher than those in PB at relapse (P=0.001).Conclusions This SP-based RT-PCR essay is a reliable method for chimerism analysis.Chimerism kinetics in BM can be used as a marker of impending leukemia relapse,especially when no other specific marker is available.Based on our findings,we recommend examining not only PB samples but also BM samples in HSCT patients.