The broader application of stem cell transplantation (SCT) for paediatric diseases has been limited by a lack of human leucocyte antigen (HLA)-matched donors. Virtually all children, however have at least one haploidentical parent who could serve as a donor. Such a donor is immediately available and the considerable costs of additional HLA typing, registry and banking expenditures that are necessary to procure an unrelated donor, could be reduced. Recent technological advances appear to have overcome the historical problems of graft rejection and severe graft versus host disease in the haploidentical setting, and in the latest studies the overall survival for children undergoing haploidentical SCT for leukaemia is now comparable with that following unrelated donor bone marrow or cord blood transplantation. Post-transplant infectious complications and leukaemia relapse remain the most important barriers yet to overcome, and new directions in the use of adoptive cellular immunity appear to be promising in this respect. Haploidentical SCT is now a viable option for those children who do not have an HLA compatible sibling or fully matched unrelated donor. The relative merits of a haploidentical family donor versus mismatched unrelated bone marrow or cord blood donation needs to be assessed in prospective, randomized clinical trials.