This chapter describes the generation of novel reagents for the treatment of cancer using fusion proteins constructed with natural ligands of the immune system. Immunotherapy is a powerful therapeutic modality that has not been fully harnessed for the treatment of cancer. We and others have hypothesized that if the proper immunoregulatory ligands can be targeted to the tumor, an effective immune response can be mounted to treat both established primary tumors and distant metastatic lesions. Though it is generally believed that immunotherapy has the potential to treat only residual disease, we offer evidence that this approach can, by itself, destroy large tumor masses and produce lasting remissions of experimental solid tumors. From these studies, three major classes of immune activators, namely, cytokines, chemokines, and costimulatory molecules, have been shown to generate antitumor responses in animal models. In addition, the reversal of immune tolerance by the deletion of T regulatory (Treg) cells has been shown to be equally important for effective immunotherapy. In an attempt to identify reagents that can provide an enhanced immune stimulation and treatment of cancer, our laboratory has developed a novel monoclonal antibody targeting approach, designated Tumor Necrosis Therapy (TNT), which utilizes stable intracellular antigens present in all cell types but which are only accessible in dead and/or dying cells. Since tumors contain necrotic and degenerating regions that account for 30-80% of the tumor mass, this targeting approach can be used to deliver therapeutic reagents to the core of tumors, a site abundant in tumor antigens. In our first set of reagents, a panel of cytokine fusion proteins was genetically engineered using monoclonal antibody chimeric TNT-3 (chTNT-3) directed against necrotic regions of tumors (single-stranded DNA) fused with IL-2, or GM-CSF, or TNFalphaa, or IFNgamma. Tested against different solid tumors, these reagents were found to mount an effective although transient immune response to tumor especially when used in combination. To improve upon these results, additional chTNT-3 fusion proteins using the liver-expression chemokine (LEC) and the costimulatory molecule B7.1 were constructed. Both of these reagents were found to work significantly better than the above cytokine fusion proteins due to their ability to stimulate multiple arms of the immune system deemed useful for cancer immunotherapy. Finally, the Tumor Necrosis Factor Superfamily (TNFSF) gene DC137L was used to generate chTNT-3 antibody (targeted) and soluble Fc (untargeted) fusion proteins. When used alone, both forms of costimulatory fusion proteins were found to produce in a s dose-dependent manner, complete regression of murine solid tumors. Evidence is presented to show that Treg cells play an important role in suppressing antitumor immunity since the deletion of these cells, when used in combination with LEC or costimulatory fusion proteins, produced profound and effective treatment with sustained memory. It is hoped that these data will further the preclinical development of soluble Fc and antibody based fusion proteins fro the immunotherapy of cancer.