The purpose of this short review is to analyse major advantages and limitations of the adenovirus (Ad), specifically with relevance to its use as a vector for gene transfer to the brain. The characteristics of Ad transduction include: the relative absence of cell type specificity; the limited spatial spread of the virus; and the long-term expression of the transgene. In the central nervous system, in contrast to that which occurs in other organs, Ad transduction in the adult does not systematically provoke cell death. Nevertheless, a proportion of the transduced cells do die, and this represents a conspicuous problem. Mechanisms leading to cell death in the brain may include immune rejection and inflammation-related toxicity, although this would not explain all of the results, and direct toxicity related to either inappropriate preparation or the transduction itself. Taking into account uncertainties concerning the innocuousness of Ad transduction, it may seem unwise to envisage Ad gene therapy for diseases that are not life-threatening and/or benefit from adequate drug or surgical treatments (e.g. Parkinson's disease or epilepsy). Ad vectors may not be easily used either in diseases displaying major immune dysfunction (e.g. multiple sclerosis). In contrast, malignant brain tumors and numerous neurodegenerative diseases (such as Huntington's, Alzheimer's diseases or amyotrophic lateral sclerosis) are directly life-threatening and deprived of any adequate treatment. They may be appropriate targets for Ad-mediated gene therapy, once both the vector and the gene of interest have been defined and optimized.