Pre-clinical studies have shown that injection of allogeneic T cells primed against a single minor histocompatibility antigen (MiHA) could cure hematologic cancers (HC) without causing any toxicity to the host. However, translation of this approach in humans has been hampered by the paucity of molecularly defined human MiHAs. Using a novel proteogenomic approach, we have analyzed cells from 13 volunteers and discovered a vast repertoire of MiHAs presented by the most common HLA haplotype in European Americans: HLA-A*02:01;B*44:03. Notably, out of >6000 MiHAs, we have identified a set of 39 MiHAs that share optimal features for immunotherapy of HCs. These 'optimal MiHAs' are coded by common alleles of genes that are preferentially expressed in hematopoietic cells. Bioinformatic modeling based on MiHA allelic frequencies showed that the 39 optimal MiHAs would enable MiHA-targeted immunotherapy of practically all HLA-A*02:01;B*44:03 patients. Further extension of this strategy to a few additional HLA haplotypes would allow treatment of almost all patients.