Recent advances in adoptive immunotherapy using cytotoxic T lymphocytes (CTLs) have led to moderate therapeutic anti-cancer effects in clinical trials. However, a critical issue, namely that CTLs collected from patients are easily exhausted during expansion culture, has yet to be solved. To address this issue, we have been developing a strategy which utilizes induced pluripotent stem cell (iPSC) technology. This strategy is based on the idea that when iPSCs are produced from antigen-specific CTLs, CTLs regenerated from such iPSCs should show the same antigen specificity as the original CTLs. Pursuing this idea, we previously succeeded in regenerating melanoma antigen MART1-specific CTLs, and more recently in producing potent CTLs expressing CD8αβ heterodimer. We are now developing a novel method by which non-T derived iPSCs are transduced with exogenous T cell receptor genes. If this method is applied to Human Leukocyte Antigen (HLA) haplotype-homozygous iPSC stock, it will be possible to prepare "off-the-shelf" T cells. As a first-in-human trial, we are planning to apply our strategy to relapsed acute myeloid leukemia patients by targeting the WT1 antigen.