Although early clinical gene therapy trials for recurrent central nervous system neoplasms showed the proof-of-principle, they did not fulfill the high expectations suggested by the preclinical experimental data. Insufficient distribution of vectors in human brain tumors and very low transduction efficiency require that we reevaluate gene transfer concepts for brain tumor treatment. Major steps to improve gene transfer into the central nervous system and the efficacy of gene therapy for malignant brain tumors include: 1) the design of more effective vector systems; 2) the development of new or improved prodrug/suicide systems, gene replacement approaches, or strategies targeting the immune response or tumor angiogenesis; 3) the study of new techniques to enhance delivery of genetic vectors into brain tumors and for monitoring gene delivery into tumors; and 4) assessment of the role of gene therapy as part of a combined treatment approach.