Adoptive transfer of antigen-specific T cells is a promising approach for preventing progressive viral infections in immunosuppressed hosts. By contrast, effective T-cell therapy of malignant disease has proven to be much more difficult to achieve. This, in part, reflects the difficulty of isolating high avidity T cells specific for tumor-associated antigens, many of which are self-antigens that have induced some level of tolerance in the host. Even when tumor-reactive T cells can be isolated, the ability of these cells to survive in vivo and traffic to tumor sites is often impaired. Additionally, most tumors employ multiple mechanisms to escape T-cell recognition, including interference in antigen presentation, secretion of inhibitory factors and recruitment of regulatory or immunosuppressive cells. The genetic modification of T cells prior to transfer provides a potential means to overcome many of these obstacles and enhance the efficacy of T-cell therapy. This review article discusses the rationale for genetic modification of T cells, the critical steps involved in gene transfer, and potential advantages and disadvantages of strategies that are now being examined to engineer improved effector T cells for the treatment of human infectious and malignant disease.