Objective: The aim of the study was to investigate the impact of the sites of high-resolution human leukocyte antigen (HLA) mismatch on the prognosis of children with leukemia undergoing umbilical cord blood transplantation (UCBT). Methods: Clinical data and high-resolution HLA-A, HLA-B, HLA-C, HLA-DRB1 and HLA-DQB1 locus gene information were collected in the children who underwent the UCBT for the first time at Children's Hospital of Soochow University between January 2016 and June 2023. In each locus, according to whether the two genes were compatible, they were divided into a compatible group (two genes were perfectly matched) and a non-compatible group (one gene was not matched). In different loci, the differences in occurrence, recurrence, non-recurrence death and survival of acute graft versus host disease (aGVHD) were compared between the two groups. Multivariate Cox regression was employed to analyzed the influencing factors for overall survival rate, and Fine-Gray proportional hazards model was employed to analyze the influencing factors of other outcome events. Results: A total of 100 patients were enrolled (55 males and 45 females), whose age [ M ( Q 1 , Q 3 )] at the time of transplantation was 3.9 (2.0, 6.5) years. There were 55 cases in the HLA-A matched group and 45 cases in the mismatched group. The 5-year non-recurrence mortality (NRM) in the HLA-A matched group was lower than that in the mismatched group ( P =0.024). The cumulative incidence of aGVHD within 100 days after transplantation in the HLA-A matched group was lower than that in the mismatched group ( P =0.017), and there were no statistically significant differences in other outcome events between the groups (all P >0.05). There were 70 cases in the HLA-B matched group and 30 cases in the mismatched group. The 5-year cumulative recurrence rate in the HLA-B matched group was higher than that in the mismatched group ( P =0.027). There were 79 cases in the HLA-C matched group and 21 cases in the mismatched group, and there were no statistically difference in the outcome events between the groups ( P >0.05). There were 73 cases in HLA-DRB1 matched group and 27 cases in mismatched group. The 5-year overall survival rate in HLA-DRB1 matched group was higher than that in mismatched group ( P =0.036), the 5-year cumulative recurrence rate in HLA-DRB1 matched group was higher than that in mismatched group ( P =0.028), and the 5-year NRM in HLA-DRB1 matched group was lower than that in mismatched group ( P =0.008). The cumulative incidence of aGVHD within 100 days after transplantation in the matched group was lower than that in the mismatched group ( P =0.010), and and there were no statistically significant difference in other outcome events between the groups ( P >0.05). There were 68 cases in HLA-DQB1 matched group and 32 cases in mismatched group. There was no statistical difference in outcome events between the two groups (all P >0.05). The risk of aGVHD in HLA-A mismatched group was higher than that in HLA-A matched group ( HR =1.25, 95% CI : 1.12-1.38). The risk of recurrence in HLA-B mismatched group was lower than that in HLA-B matched group ( HR =0.77, 95% CI : 0.63-0.91). Mismatched group at HLA-DRB1 compared with matched group at HLA-DRB1, had a higher risk of aGVHD ( HR =1.37, 95% CI : 1.26-1.48), a higher risk of non-recurrence death ( HR =1.39, 95% CI: 1.28-1.50), and a higher risk of death ( HR =1.27, 95% CI : 1.18-1.36). No association was found between HLA-C and HLA-DQB1 locus with the risk of aGVHD, recurrence, non-recurrence death, and survival (all P >0.05). Conclusions: In UCBT, the risk of aGVHD in children with matching HLA-A sites of donor and recipient is lower than that in children with incompatible HLA-A sites. Compared with children with incompatible HLA-DRB1 sites, children with HLA-DRB1 matched sites has a lower risk of acute GVHD, a lower 5-year NRM, and a higher risk of death. The recurrence rate of children with matching HLA-B loci is higher than that of children without matching HLA-B loci.