BCR-ABL+ acute lymphoblastic leukemia patients have transient responses to current therapies. However, the fusion of BCR to ABL generates a potential leukemia-specific Ag that could be a target for immunotherapy. We demonstrate that the immune system can limit BCR-ABL+ leukemia progression although ultimately this immune response fails. To address how BCR-ABL+ leukemia escapes immune surveillance, we developed a peptide: MHC class II tetramer that labels endogenous BCR-ABL–specific CD4+ T cells. Naive mice harbored a small population of BCR-ABL–specific T cells that proliferated modestly upon immunization. The small number of naive BCR-ABL–specific T cells was due to negative selection in the thymus, which depleted BCR-ABL–specific T cells. Consistent with this observation, we saw that BCR-ABL–specific T cells were cross-reactive with an endogenous peptide derived from ABL. Despite this cross-reactivity, the remaining population of BCR-ABL reactive T cells proliferated upon immunization with the BCR-ABL fusion peptide and adjuvant. In response to BCR-ABL+ leukemia, BCR-ABL–specific T cells proliferated and converted into regulatory T (Treg) cells, a process that was dependent on cross-reactivity with self-antigen, TGF-β1, and MHC class II Ag presentation by leukemic cells. Treg cells were critical for leukemia progression in C57BL/6 mice, as transient Treg cell ablation led to extended survival of leukemic mice. Thus, BCR-ABL+ leukemia actively suppresses antileukemia immune responses by converting cross-reactive leukemia-specific T cells into Treg cells.