The potential for immune-mediated destruction of neoplasms was suggested nearly one century ago. Despite this, no "magic bullet" has yet been identified. Nevertheless, the physiology of cell-mediated immune reactions has been well characterized in molecular, cellular, and clinical studies of allograft and microbial immunity. Extensive studies performed in laboratory animal models have documented the in vitro and in vivo destruction of various neoplastic tissues by immune cells. This destruction can be directed against autologous, syngeneic, or allogeneic tumors in several systems with varying degrees of "tumor specificity". Two approaches exist towards utilizing these immune reaction in vivo. The first involves providing the tumor bearer with immunostimulatory agents, either specific or nonspecific, designed to activate and amplify the destructive potential of the individual's endogenous immune cells able to recognize and destroy autologous tumor. The second approach provides immune cells with antitumor capacity to a tumor-bearing individual, these cells having been activated exogenously. A number of successful regimens involving these two approaches, and combinations of them, have been delineated in animal tumor models. These experimental studies lay a strong foundation for initiating clinical trials of these concepts for patients with cancer. This review summarizes the diverse experimental studies in animals leading to clinical trials, presents recent data from ongoing clinical trials directly testing the potential for cellular immunotherapy, and then presents some of the major challenges facing further development and application of this potential therapeutic approach.