Gliomas affect 15,000 to 17,000 Americans every year and carry a dismal prognosis. The potential of immunologically mediated diagnosis and therapy, although greatly enhanced since the advent of monoclonal antibodies, has not been fully realized due to significant problems, most especially the challenge of identifying antigenic molecules specific to glial tumors. Other problematic issues include antigen-associated factors such as heterogeneity, modulation, shedding, and cross-reactivity with normal cells, and factors associated with therapeutic agent delivery, typically variable tumor perfusion and unfavorable diffusional forces in tumor microenvironment. An understanding of these problems called for the delineation of operationally specific antigens (tumor-associated antigens not expressed by the normal central nervous system) combined with the use of compartmental therapeutic approaches to increase the specificity of therapy. Numerous antigens have been identified and are classified as extracellular/matrix-associated, membrane-associated, and intracellular antigens. Nevertheless, only a few have been demonstrated to be of significant therapeutic and diagnostic utility. These few include the extracellular matrix-associated antigens tenascin and GP 240, defined by the antibody DMAb-22. Recent identification of the overexpression of a deletion variant of the epidermal growth factor receptor (EGFRvIII) in up to 50% of the more malignant glial tumors and the subsequent creation of monoclonal antibodies that are specific to this molecule and do not recognize the wild-type EGFR provide the most exciting development yet in the design of specific antiglioma immunoconjugates. In addition, the tumor-specific nature of EGFRvIII combined with improved knowledge of immune mechanisms, especially in the context of the central nervous system, will facilitate the design of highly selective cell-mediated therapeutic approaches with a view toward obtaining tumor-specific immunity. © 1995 Wiley-Liss, Inc.