Adoptive transfer of antigen-specific T lymphocytes is an attractive form of immunotherapy for haematological malignancies and cancer. The difficulty of isolating antigen-specific T lymphocytes for individual patients limits the more widespread use of adoptive T cell therapy. The demonstration that cloned T cell receptor (TCR) genes can be used to produce T lymphocyte populations of desired specificity offers new opportunities for antigen-specific T cell therapy. The first trial in humans demonstrated that TCR gene-modified T cells persisted for an extended time period and reduced tumor burden in some patients. The WT1 protein is an attractive target for immunotherapy of leukemia and solid cancer since elevated expression has been demonstrated in AML, CML, MDS and in breast, colon and ovarian cancer. In the past, we have isolated high avidity CTL specific for a WT1-derived peptide presented by HLA-A2 and cloned the TCR alpha and beta genes of a WT1-specific CTL line. The genes were inserted into retroviral vectors for transduction of human peripheral blood T lymphocytes of leukemia patients and normal donors. The treatment of leukemia-bearing NOD/SCID mice with T cells transduced with the WT1-specific TCR eliminated leukemia cells in the bone marrow of most mice, while treatment with T cells transduced with a TCR of irrelevant specificity did not diminish the leukemia burden. In order to improve the safety and efficacy of TCR gene therapy, we have developed lentiviral TCR gene transfer. In addition, we employed strategies to enhance TCR expression while avoiding TCR mis-pairing. It may be possible to generate dominant TCR constructs that can suppress the expression of the endogenous TCR on the surface of transduced T cells. The development of new TCR gene constructs holds great promise for the safe and effective delivery of TCR gene therapy for the treatment of malignancies.