ABO incompatibility does not hinder Bone Marrow Transplantation (BMT), but it has been associated with worse outcomes and additional adverse events. Many studies described effects of ABO incompatibility on allogeneic BMT, including delayed engraftment of Red Blood Cells (RBC), platelets and/or neutrophils; Pure Red Cell aplasia (PRCA); helolytic reaction; reduced overall survival, beside others. However, there is no standard recommendation regarding the maximum safe value of RBC allowed in an ABO-incompatible transplant. We compared major and bidirectional ABO incompatible (iABO) BMT (n = 42) with ABO compatible BMT (n = 44) and evaluated the impact of the value of incompatible Red Blood Cells (iRBC) in outcomes and adverse events, to determine a safe iRBC value to be infused. The impact of infused RBC was evaluated using two dose parameters: volume of infused RBC/Kg and number of infused RBC/Kg. Characteristics regarding patients/transplantation of ABO-compatible and -incompatible were similar in age, weight, donor type, disease, conditioning regimen, dose of TNC/Kg and dose of CD34+ cells/Kg. The iABO patients presented a delayed time of transfusion independence in 30/60 days, increased requirement for RBC transfusion, more hemolysis signals, and incidence of PRCA. Neutrophils/platelets engraftment, length of hospitalization post-transplant, platelet units requirement, graft-versus-host-disease occurrence, and overall survival were similar in both groups. Patients of the iABO group received 0.93 mL RBC/Kg (range 0.31–3.50) and 1.03×1010 RBC/Kg (range 0.36–3.88). The infusion of Irbc > 1.0 mL/Kg or > 1.0 ×1010 RBC/Kg was related with graft failure or death before neutrophil and/or platelet engraftment, increased fresh frozen plasma requirement, and increased creatinine. Our results also suggest that antibodies titers impact the transplantation scenario. In this study, the high antibodies titers were related with hemolysis, PRCA, delayed RBC engraftment and adverse reactions Grade at infusion. Additionally, we propose the use of the number of RBC (RBC/Kg) as a more appropriate dose parameter of infused iRBC. Comparing volume of RBC/Kg and number of RBC/Kg, the Spearman correlation was 0.994 and the outcomes were similar when used one or other parameter. The number of RBC is a direct measure of hemocytometer and disregards the size of cells, while the volume of RBC is estimated using a relative parameter of hemocytometer. Moreover, all other dose cell-related parameters are used in function of the number of cells. Finally, further studies involving larger cohorts, with a larger range of infused RBC values, and addressing more factors, are necessary to clarify the maximum safe value of RBC to be infused in iABO transplants.