Our goal was to convert murine RM-9 prostate carcinoma cells in vivo into antigen-presenting cells capable of presenting endogenous tumor antigens and triggering a potent T-helper cell-mediated immune response essential for the generation of a specific antitumor response. We showed that generating the major histocompatibility complex (MHC) class I+/class II+/Ii- phenotype, within an established subcutaneous RM-9 tumor nodule, led to an effective immune response limiting tumor growth. This phenotype was created by intratumoral injection of plasmid cDNAs coding for interferon gamma, MHC class II transactivator, and an antisense reverse gene construct (RGC) for a segment of the gene for Ii protein (-92,97). While this protocol led to significant suppression of tumor growth, there were no disease-free survivors. Nevertheless, irradiation of the tumor nodule on the day preceding initiation of gene therapy yielded 7 of 16 mice that were disease-free in a long-term follow up of 57 days compared to 1 of 7 mice receiving radiotherapy alone. Mice receiving radiotherapy and gene therapy rejected challenge with parental RM-9 cells and demonstrated specific cytotoxic T-cell activity in their splenocytes but not the mouse cured by radiation alone. These data were reproduced in additional experiments and confirmed that tumor irradiation prior to gene therapy resulted in complete tumor regression and specific tumor immunity in more than 50% of the mice. Increasing the number of plasmid injections after tumor irradiation induced tumor regression in 70% of the mice. Administering radiation before this novel gene therapy approach, that creates an in situ tumor vaccine, holds promise for the treatment of human prostate carcinoma.