Among the broad array of genes that have been evaluated for tumor therapy, those encoding prodrug activation enzymes are especially appealing as they directly complement ongoing clinical chemotherapeutic regimes. These enzymes can activate prodrugs that have low inherent toxicity using both bacterial and yeast enzymes, or enhance prodrug activation by mammalian enzymes. The general advantage of the former is the large therapeutic index that can be achieved, and of the latter, the non-immunogenicity (supporting longer periods of prodrug activation) and the fact that the prodrugs will continue to have some efficacy after transgene expression is extinguished. This review article describes 13 different prodrug activation schemes developed over the last 15 years, two of which - activation of ganciclovir by viral thymidine kinase and activation of 5-fluorocytosine to 5-fluorouracil - are currently being evaluated in clinical trials. Essentially all of these prodrug activation enzymes mediate toxicity through disruption of DNA replication, which occurs at differentially high rates in tumor cells compared with most normal cells. In cancer gene therapy, vectors target delivery of therapeutic genes to tumor cells, in contrast to the use of antibodies in antibody-directed prodrug therapy. Vector targeting is usually effected by direct injection into the tumor mass or surrounding tissues, but the efficiency of gene delivery is usually low. Thus it is important that the activated drug is able to act on non-transduced tumor cells. This bystander effect may require cell-to-cell contact or be mediated by facilitated diffusion or extracellular activation to target neighboring tumor cells. Effects at distant sites are believed to be mediated by the immune system, which can be mobilized to recognize tumor antigens by prodrug-activated gene therapy. Prodrug activation schemes can be combined with each other and with other treatments, such as radiation, in a synergistic manner. Use of prodrug wafers for intratumoral drug activation and selective permeabilization of the tumor vasculature to prodrugs and vectors should further increase the value of this new therapeutic modality.