Tumor cells often evoke specific immune responses that, however, fail to eliminate all the tumor cells. The development of cancer immunotherapies is, therefore, mostly focused on the generation of large numbers of activated anti-tumor effector cells by vaccination or adoptive T cell transfer. These developments are built on an ever-extended list of identified tumor-associated antigens and corresponding T cell epitopes, and a steady flow of reports from proof-of-principle animal model experiments demonstrating cure from disease by immune interventions. However, the promises have not translated into clinical successes for cancer patients. Even where tumor regression or complete responses were achieved there is usually relapse of the disease. Increasing numbers of reports over recent years highlight potential immunosuppressive mechanisms that act in tumors and systemically in cancer patients to block effective anti-tumor immune responses. They account in large parts for the failures of cancer immunotherapy and need to be overcome before progress can be expected. We review here the current state of the research on immunosuppressive networks in human cancer.