The possibilities of immunolog&al intervention in neoplastic diseases are discussed in light of the accumulating findings that many spontaneously arising tumors of animals are not immunogenic in the host of origin or are not susceptible to immunological attack, and of the doubts as to the existence of tumor-associated antigens for many human neoplasms. Although some spontanous tumors do in fact display immunogenic properties and are subject to cytotox[c immune reactions, there are persuasive arguments for anticipating that a large proportion of naturally appearing malignant growths will be found to lack both immunogenicity and sensitivity to immunological effector mechanisms in the course of the host-tumor association. Even in the case of such tumors, however, significant opportunities for therapy and prophylaxis by extrinsic immunological manipulation remain cogent, as do eventualities of immunodiagnosis. These possibilities are discussed. It is concluded that whereas the simplistic assumptions which for some years set expectations in the field of tumor immunology must now be abandoned, reassessment of the nature of immunological parameters of host-tumor relationships and of the scope of immunological intervention on the basis of the information now available suggests that the immunological approach to the control of cancer continues to be a profitable area of investigation.