Background:Immune hemolytic disease (IHD) is a main and severe complication after hematopoietic stem cell transplantation (HSCT). Allogeneic immune‐hemolysis is that immune cells and antibodies from the recipient attack red blood cells from donor, which is defined as host versus graft (HVG). We analyzed retrospectively our data and given a summary as below.Aims:To try to illustrate and proven the causes of Post‐HSCT IHD, which is GVG or HVG, by analysis based on serum titer of ABO antibody, chimerism rate of donor cell, and responds of therapy.Methods:707 HSCTs from our center since January 2010 to March 2017, were retrospectively analyzed. The item of analysis included age, gender, ABO blood‐type, diseases and transplant type, etc. Standard of diagnosis of Post‐HSCT IHD included: 1: after HSCT, 2:hemoglobin ≤80 × 109/L for three time in two consecutive weeks, 3: needed regular transfusion, and 4 positive Coomb's test. Standard of partial remission: without gaining CR standard but ≥ 50% decline of transfusion. Statistic methods included stata13 chi‐square test.35 cases were divided into three groups by host blood‐type antibody titer and donor cell chimerism. Group 1(n = 16), all cases were donor‐recipient ABO blood‐type mismatched and 100% donor cells chimerism but recipient ABO blood‐type antibodies occurrence: antibody titer high (more than pre‐HSCT value) or delayed decline (more than 3 months after HSCT). Group 2 (n = 13), mixed donor cell chimerism (75–98.8%). Group 3 (n = 6) full donor cell engraftment, 100% donor cell chimerism without ABO blood‐type antibody from recipient. In group 3, four with donor‐recipient ABO mismatch, 3 cases with acute GVHD and 5 cases with positive CMV test. In G1, ABO blood‐type antibodies from recipient can indirectly prove host immune cell existence. In G2, chimerism <100% may directly prove host immune cell existence. In G3, although we had no direct evidence to prove recipient cells existence and some cases were along with GVHD, by responds of therapy host cells were still thought to be existence according to hypothesis of immune cells escapes and immune cells harbor in lymph nodes.Results:Post‐HSCT IHD incidence was 4.95%, 336 patients with ABO blood‐type mismatch, 27 suffered from post‐HSCT IHD, In contrast, oppositely only 8 patients had post‐HSCT IHD among 339 patients with ABO blood‐type match (7.14% versus 2.36%, p = 0.001). Post‐HSCT IHD occurred in 10 cases in 40 HSCT pairs of A blood‐type donor and O blood‐type recipient (A‐O pair), 2 in 54 B‐O pairs and 2 in 10 AB‐O pairs. statistic significant difference was found when comparing three groups (p = 0.029). According to Chi‐square test, Mixed chimerism compared with 100% implantation, in adverse implantation the probability of immuno‐hemolysis was greater, it was statistically significant(P = 0.000, P < 0.05). The percentage of hemolysis after transplantation for patients with matched sibling donors was 3.44%, for patients with unrelated donors was 7.38%(P = 0.019<0.05), which appeared that the probability of post‐transplantation hemolysis by unrelated donors was greater.Summary/Conclusion:The cells and antibody leading post‐HSCT IHD from host cells were proved by all 35 patients with post‐HSCT IHD, who was cured with tapering immunosuppress drugs, plus or not Rituximab, and giving or not DLI. High titer recipient blood‐type antibody was a direct evidence of existence of host immune cells. Tapering immunosuppress drugs, plus or not Rituximab and giving or not DLI were a correct therapy for post‐HSCT IHD.