Goals Chimerism analysis (CA) is essential for post-transplant surveillance. DNA from bone marrow (BM) aspirates drawn into EDTA specimen tubes (EDTA-BM) is widely used for CA. Since EDTA-BM is subject to peripheral dilution, however, DNA from aspirate particle smears (PS-BM) might better represent BM chimerism status. In this regard, we evaluated the influence of BM sources on CA. Procedures Study subjects were consecutive pediatric AML patients who had experienced relapse after allogeneic stem cell transplantation with the interval between CA before relapse (pre-relapse) and at relapse <6 months. We compared chimerism status at the 2 time-points by fluorescence PCR on STR markers using EDTA-BM vs PS-BM. Results Eight patients were eligible for this study. The recipient DNA (%R) from EDTA-BM was 0% at pre-relapse in all except 2 with 1.6% and 1.3%, while %R using PS-BM revealed mixed chimerism in all 8, %R ranging 1.6-13.2% (median 3.75%). The %R from EDTA-BM was 0.9-79.3% at relapse (29.15%), while %R from PS-BM was 3.8-86.6% (60.15%). The difference of %R (Δ%R), %R[PS-BM]-% R[EDTA-BM], was median 3.15% (1.6-13.2%) at pre-relapse and median 12.1% (2.1-60.7%) at relapse. Conclusions Our study showed CA using EDTA-BM significantly underestimated %R. Our observation might have a ramification to other quantitative workup in hematologic malignancy.