A central paradox in tumor immunology is the seeming coexistence of a developing or progressing tumor burden in the absence of a therapeutically effective host anti-neoplastic response. It is now generally thought, however, that the host, whether mouse or human, can mount a T cell response against a developing malignancy without any prior immunization. Ostensibly, the nature or potency of that antitumor response is frequently insufficient to alter the course of neoplastic progression. Consequently, cancer immunotherapy was born from the idea that a preexisting or newly induced immune response may be intensified to achieve clinical regression. At the heart of that matter is the capacity of the host immune system to not only phenotypically distinguish normal from malignant cells, but also to functionally mediate antitumor activity. In both preclinical models and clinical settings, a plethora of tumor-associated or tumor-specific antigens have now been identified that can induce anti-neoplastic CD4+ or CD8+ T cell responses. Thus, immunization against neoplastic disease is attainable and, in several cases, overcomes potential mechanisms of immune tolerance and escape. In terms of objective clinical responses, however, the results have been less dramatic, and only a smaller proportion of patients have undergone significant disease regression. These responses have been observed mainly in clinical settings of adoptive T cell therapy, as compared with vaccination approaches. Nonetheless, these studies collectively provide the rationale for the further development and application of active and adoptive immunotherapies, perhaps in combination with each other or with oncological interventions, to promote more meaningful and sustainable anti-neoplastic immune responses and clinical outcomes.