By far, cancer accounts for the majority of gene therapy trials that are being carried out worldwide. Seventy percent of the gene therapy protocols that have been reviewed by the National Institutes of Health Recombinant DNA Advisory Committee (NIH RAC) are for the treatment of cancer. Of these, two thirds involve immunotherapy, with transfer of genes for cytokines, immune accessory molecules, or tumor antigens into a variety of cellular targets. Other clinical protocols include chemoprotection, prodrug activation, or tumor-suppressor gene replacement. Either local or distal bystander effects may mediate antitumor effects. These bystander mechanisms may help to overcome poor transduction efficiencies by currently available vectors. Replicating oncolytic viruses entering the clinic include herpes virus, Newcastle disease virus, reovirus, and others. These viruses have been shown to replicate selectively in cancer cells, albeit by different mechanisms. Reovirus, for example, requires the presence of an activated Ras signaling pathway in order to replicate and destroy cells. Tumor selectivity can be achieved by placing an essential viral gene under the control of a tumor-specific promoter. The tumoricidal effects of replicating viruses may be enhanced by genetic modification-for example, by the insertion of a cytokine gene to elicit antitumor immunity. Clearly, much work needs to be done both in the laboratory and in the clinic in order to exploit the full potential of these novel gene and viral therapies.