It has been known for many years that cell-mediated immune mechanisms are primarily responsible for rejection of normal tissue grafts in genetically dissimilar individuals, and it is now clear that immunity to cancer is mediated in the same way. Malignant tumors have cell surface antigens which are different from those of the tissues from which they arise, hence the term “tumor-specific antigens.” Largely responsible for this discovery was the development of syngeneic (genetically identical) host systems which have permitted assessment of tumor-specific immunity rather than the transplantation antigens of allograft immunity. Host-tumor surveillance is primarily lymphocyte mediated. The humoral immune mechanism is also involved, but the best documented action of humoral factors is interference with cell-mediated tumor destruction by “blocking.” Direct evidence for a host-tumor relationship has led to a resurgence of interest in immunologic therapy for cancer. Prospects for immunotherapy include nonspecific active immunization (i.e., BCG), transfer of lymphocytes or transfer factor, selective elimination or inactivation of humoral blocking material, and prolongation of tumor-specific antigen stimulation by in situ necrosis. At the present time it is unlikely that immunotherapy alone will play a major role in the treatment of gross deposits of cancer, but it may become an important adjunct to surgical therapy by eradicating small populations of tumor cells, both within the surgical field and at distant sites. Theoretically, immunotherapy could solve both of these problems which are the consequences of small undetectable inocula of tumor cells having been deposited at distant sites or in the surgical field. Before immunotherapy can be accepted as a standard form of cancer therapy, it will be necessary to delineate the interrelationships of tumor antigens, cytotoxic antibodies, blocking factors, and delayed type hypersensitivity.