Malignant primary and metastatic brain tumours continue to be associated with poor prognosis. Nevertheless, recent advances in molecular medicine, specifically in the strategies of gene therapy, targeting tumour cells, anti-angiogenesis and immunotherapy, have created novel tools that may be of therapeutic value. To date, gene therapy trials have not yet demonstrated clinical efficacy because of inherent defects in vector design. Despite this, advances in adenoviral technology, namely the helper-dependent adenoviral constructs (gutless) and the uncovering of brain parenchymal cells as effective and necessary targets for antitumour benefits of adenoviral-mediated gene transfer, suggest that developments in vector design may be approaching the point of clinical utility. Targeting tumour cells refers to strategies that destroy malignant but spare normal cells. A new assortment of oncolytic viruses have emerged, capable of specific lysis of cancer tissue while sparing normal cells and propagating until they reach the tumour borders. Furthermore, peptides have been transformed into bullets that specifically seek and destroy cancer cells. The concept of tumour angiogenesis has been challenged by new but still very controversial findings that tumour cells themselves may form blood channels. These results may lead to the redirecting of the molecular targets toward anti-angiogenesis in some tumours including glioblastoma multiform. Unfortunately, our knowledge regarding the immunological ignorance of the tumour is still limited. Even so, newly discovered molecules have shed light on novel pathways leading to the escape of the tumour from the immune system. Finally, significant limitations in our current experimental tumour models may soon be overcome by firstly, the development of models of reproducible organ-specific tumours in non-inbred animals and secondly applying genomics to individualize therapy for a particular tumour in a specific patient.