Although many patients with malignant brain tumors can be rendered free of bulk disease by current surgery and radiotherapy techniques, complete tumor eradication is extremely difficult to achieve, raising interest for T-cell adoptive immunotherapy. Conclusive experimental data generated by many investigators coupled with clinical experience have debunked many of the theoretical 'obstacles' to immunotherapy in the CNS. First, there does not appear to be a significant vascular barrier in brain tumors to prevent trafficking of systemically administered activated T cells. Moreover, T cells stimulated in the periphery by DC vaccination are able to mediate regression of established intracranial tumors. Second, brain tumor patients are able to mount an immune response against autologous tumor. Not surprisingly, patient factors such as tumor burden, corticosteroid use, advanced age, or recent chemotherapy can inhibit the immune response to tumor. Directing this type of therapy to patients without these factors may improve the likelihood of response. Third, therapeutic immune responses occurring within the CNS against tumors derived from CNS tissue have not been associated with clinical signs of autoimmune reactions against normal brain tissue. The general toxicity associated with the systemic adoptive transfer of ex vivo activated LN cells is very low. Toxicity of activated T cells delivered locally into the tumor resection cavity likewise is low and transient. Current phase II clinical trials of AI are in progress to determine the response rate for patients with newly diagnosed malignant gliomas. Future developments to characterize shared brain tumor antigens and develop more effective strategies for vaccination may lead to a more effective and broadly applicable therapy for CNS malignancies.