The discovery of a large array of tumor antigens has demonstrated that host lymphocytes can indeed recognize and destroy tumor cells as originally proposed in the cancer immunosurveillance hypothesis. Recent reports that led to the cancer immunoediting concept also strongly support the immunosurveillance hypothesis, and they further indicate that the host immune system plays a critical role not only in promoting host protection against cancer but also in selecting tumors that can better escape from immune attack. Thus, it is now clear that T cells have the ability to recognize and destroy spontaneously arising tumors. However, the generation of antitumor immunity is often difficult in the tumor-bearing host because of various negative regulatory mechanisms. Here, we review our recent work on tumor immunotherapy, which utilizes the activation of type-1 innate and/or acquired immunity as a potent strategy to overcome immunosuppression in the tumor-bearing host. We have established a variety of tumor therapeutic protocols that aim to activate type-1 immunity, such as tumor-vaccine therapy with CpG encapsulated in liposomes, cell therapy using tumor-specific Th1 cells, and gene therapy using gene-engineered Th1 cells. We found that CpG encapsulated in liposomes stimulated IL-12-producing DC and induced IFN-γ γ γ γ-producing NK cells, NKT cells, and tumorspecific CTL. Th1 cell therapy was also shown to be beneficial for acceleration of APC/Th1 cell-cell interaction in the DLN, which was critical for inducing tumor-specific CTL at the tumor site. Therefore, we conclude that the activation of type-1 innate and acquired immunity is crucial for tumor immunotherapy in order to overcome strong immunosuppression in the tumor-bearing host. (Cancer Sci 2004; 95: 697-703) ince the cancer immunosurveillance hypothesis was proposed by Ehrlich, Burnet, and Thomas, 1, 2) tumor immunologists have asked whether the host immune system can prevent tumor growth. Recent reports from several groups have re-addressed this issue and provided strong evidence for the existence of an effective cancer immunosurveillance process in mice.3-11) As summarized in Table 1, mice lacking specific cellular populations, such as T cells, natural killer T (NKT) cells, and/or natural killer (NK) cells, as well as specific molecules, such as interferon (IFN)-γ, interleukin (IL)-12, perforin, or tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), unequivocally show higher incidences of tumor development. Moreover, administration of IL-12 or α-galactosylceramide (α-GalCer) that can stimulate T, NKT, and/or NK cells, reduces primary tumor incidence.Although it is much more difficult to address cancer immunosurveillance in humans, three lines of evidence suggest that cancer immunosurveillance indeed occurs in humans: (i) immunosuppressed transplant recipients display higher incidences of non-viral tumors, such as melanomas, colon, lung, pancreas, bladder, kidney, and endocrine system cancers, than agematched immunocompetent control populations, ...