The introduction of post-transplantation cyclophosphamide (PTCy) as graft-versus-host disease (GVHD) prophylaxis has made haploidentical (haplo) hematopoietic stem cell transplantation (HSCT) a common approach in adults, but pediatric experience is limited. Based on the encouraging adult data and with the aim of decreasing the risk of graft failure, our center is increasingly using peripheral blood stem cells (PBSCs) from haplo donors with PTCy. Here we compare outcomes of bone marrow (BM) transplantation with traditional donor choices, including matched sibling donors (MSDs) and 10/10 HLA matched unrelated donors (MUDs), with those of haplo PBSC grafts in pediatric patients with hematologic malignancies. In this retrospective single-center study, the primary endpoint was the comparison of GVHD-free relapse-free survival (GRFS; defined as absence of acute GVHD [aGVHD] grade III-IV, relapse, death, or chronic GVHD [cGVHD] requiring systemic therapy) for the 3 cohorts. Secondary endpoints included overall survival (OS), relapse-free survival (RFS), nonrelapse mortality (NRM), and incidence of aGVHD and cGVHD). A total of 104 consecutive patients underwent first allogeneic (allo)-HSCT for a hematologic malignancy or myelodysplastic syndrome between January 2014 and December 2020 using a haplo family donor (PBSCs; n = 26), an MSD (BM; n = 31), or an MUD (BM; n = 47). Patient demographic and transplantation characteristics were not significantly different across the cohorts, apart from remission status, with the haplo cohort having more patients in third or later complete remission before HSCT (P < .01). The median duration of follow-up for the entire cohort was 573 days. The cumulative incidence of aGVHD (grade II-IV or grade III-IV) was not significantly different among the cohorts; however, the cumulative incidence of cGVHD at 18 months was highest in the MUD cohort (31.7%, versus 10.0% in the MSD cohort and 9.2% in the haplo cohort; P = .02). There were no differences in the 18-month cumulative incidence of relapse or NRM. OS and RFS at 18 months were 80.7% (95% confidence interval [CI], 61.7% to 100%) and 73.8% (95% CI, 55.5% to 98.1%) for the haplo cohort, 83.4% (95% CI, 72.8% to 95.5%) and 70.3% (95% CI, 57.9% to 85.3%) for the MUD cohort, and 80.9% (95% CI, 66.9% to 97.7%) and 66.5% (95% CI, 50.5% to 87.5%) for the MSD cohort, with no statistically significant differences among the cohorts. GRFS at 18 months was 61% (95% CI, 43.3% to 85.9%) for the haplo cohort, 44.6% (95% CI, 31.8% to 62.5%) for the MUD cohort, and 62.1% (95% CI, 45.7% to 84.3%) for the MSD cohort (P = .26). Haploidentical PBSC HSCT with PTCy had comparable outcomes to MSD and MUD BM HSCT and less cGVHD compared with MUD BM HSCT in children. The logistical advantages and lower resource burden of haplo HSCT with PBSCs make it a feasible alternative to MUD HSCT in children with hematologic malignancies. Given that this is a retrospective comparison of transplantation platforms rather than donor types, further prospective studies are warranted. © 2021 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc., (Copyright © 2021 The American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc. All rights reserved.)