Hematopoietic cell transplantation (HCT) can be a life-saving therapy for patients with genetic and acquired hematologic diseases. Despite major advances in supportive care during HCT, immunological complications of the alloimmune response, including graft rejection and graft-versus-host disease (GVHD), remain major impediments to successful clinical outcomes. Although graft rejection mediated by host immune cells and GVHD mediated by donor immune cells can be prevented or mediated by immune suppression therapy, genetic HLA matching remains essential for successful strategies designed to minimize the risks of transplantation. The most favorable HCT results are seen in patients with a genotypically HLA-identical sibling donor, but the limited availability of matched related donors has severely restricted the clinical application of this therapy. Fortunately, the establishment of large unrelated volunteer donor registries now provides the opportunity to identify HLA matches for many patients who lack a family donor. The criteria for unrelated donor matching, however, are poorly defined. Until recently, an analysis of matching beyond HLA-identical siblings has been limited by typing technology. The introduction within the past few years of new methods for high resolution typing and definition of HLA alleles has had a profound impact on our ability to identify and interpret the multiple nucleotide sequence polymorphisms that encode HLA antigens. Preliminary studies clearly demonstrate the importance of precise matching at the allele level for successful transplantation. There remain, however, important unanswered questions about the relative importance of different HLA loci in matching strategies, as well as incomplete information about permissible limits of mismatching in different patient populations.