Research on human tumor immunology has greatly advanced in the past two decades. Many immunogenic tumor antigens have been identified, and some of these antigens entered in clinical trials. Consequently, it has been shown that these antigens can inhibit tumor growth in patients to some extent, indicating that they act as potent immunogenic therapeutic vaccines in cancer patients with malignancies originating from various tissues. These patients had antigen-specific cytotoxic T-lymphocyte (CTL) responses when assessed on tetramer, enzyme-linked immunospot (ELISPOT), T-cell clonotype and CTL induction efficiency. Thus, it has become clear that human tumor vaccines can evoke clinical and immunological anti-tumor responses in patients. The tumor regression effects of tumor vaccines, however, are generally low, and it is obvious that current vaccination protocols are generally too weak to provide substantial and satisfactory clinical benefits. This means that other drastic and more potent clinical and immunological protocols are required in cancer immunotherapy. To find such efficient protocols the basic immunological and biological properties of cancers must be investigated. In the present review the identification of human tumor antigens recognized on CTL and the clinical trials are introduced. Next, the most recent analysis of human cancer-initiating cell (cancer stem cell)-associated antigens is described. These antigens might be able to act as 'universal, general and fundamental' tumor antigens. Also present is the authors' recent study for increasing cross-presentation efficiency in dendritic cells and subsequent enhancement of human leukocyte antigen (HLA)-class I-restricted peptide antigenicity by using HSP90 and ORP150 molecular chaperones that act as endogenous Toll-like receptor ligands. In addition to the aforementioned manipulation of the positive loop of tumor immunity, it is necessary to regulate and intervene in the negative loop. In particular, the potential of the expression of HLA class I molecule regulation by epigenetic mechanisms will be discussed. Finally, the type of basic and clinical tumor immunology research highly required currently, and in the very near future, are described.