Retrovirus (RV) has been one of the earliest recombinant vectors to be investigated in the context of cancer gene therapy. Experiments in cell culture and in animal brain tumor models have demonstrated the feasibility of RV mediated gene transduction and killing of glioma cells by toxicity generating transgenes. Phase I and II clinical studies in patients with recurrent malignant glioma have shown a favorable safety profile and some efficacy of RV mediated gene therapy. On the other hand, a prospective randomized phase III clinical study of RV gene therapy in primary malignant glioma failed to demonstrate significant extension of the progression-free or overall survival times in RV treated patients. The failure of this RV gene therapy study may be due to the low tumor cell transduction rate observed in vivo. The biological effects of the treatment may also heavily depend on the choice of transgene/prodrug system and on the vector delivery methods. Retrovirus clinical trials in malignant glioma have nevertheless produced a substantial amount of data and have contributed toward the identification of serious shortcomings of the non-replicating virus vector gene therapy strategy. Novel types of therapeutic virus vector systems are currently being designed and new clinical protocols are being created based on the lessons learned from the RV gene therapy trials in patients with malignant brain tumors.