In hereditary neurological diseases, gene transfer into neurons is made difficult by: the nature of the cells (postmitotic cells, that cannot be cultured, genetically modified ex vivo, then retransplanted), sometimes, their widespread localization, the blood-brain barrier. However, three viral vectors derived from adenovirus, Herpes simplex virus and adeno-associated virus have been shown to be very efficient in transferring DNA into brain cells. All of these vectors can infect resting cells, especially neurons, and are efficient in vivo. Retroviral vectors which can infect dividing cells only are mainly used for ex vivo genetic modification of cells (neural progenitor cells, myoblasts, fibroblasts) followed by intracerebral transplantation. Alternatively, genetically modified cells can be transplanted in a peripheral site if the transgene product is able to cross the blood-brain barrier or to be transported retrogradely from the nerve terminals. We have especially investigated the potential interest of adenoviral vectors to transfer foreign genes into brain cells and to treat animal models of neurological diseases. These vectors allowed us to transfer the lacZ gene into any neural cell type, including neurons, glia, photoreceptors and olfactory receptors, ex vivo, in cell culture, and in vivo, by stereotactic administration. In addition, axonal transport of adenoviral vectors has been demonstrated, e.g. in the substantia nigra after injection into the striatum, in the olfactory bulb after intranasal instillation and in spinal motor neurons after intramuscular injection. After intracerebroventricular injection, ependymal cells are massively infected and express the transgene for several months, as this is also observed in neurons. Through the spinal canal and cerebrospinal fluid, the vector can diffuse to a considerable distance from the injection point, e.g. to the lumbar spinal cord after injection in the suboccipital region. To test the biological function of transgenes transferred through adenoviral vectors, we have constructed vectors with cDNAs or genes for various neutrophic factors: CNTF, NT3, BDNF and GDNF. These vectors were biologically active on target cells, ex vivo and in vivo. In the pmn mouse model of progressive motor neuronal degeneration, some of these vectors, alone or combined, allowed for prolongation of life of homozygous animals by more than two fold, and for decrease in the demyelination of phrenic nerve axons. Finally, we have also constructed an adenoviral vector carrying the alpha-hexosaminidase cDNA, encoding the enzyme subunit deficient in Tay Sachs patients. This vector permitted to normalize ganglioside metabolism in Tay Sachs fibroblasts and is currently tested in knock out mice deficient in hexosaminidase A. In spite of all these encouraging results, we are nevertheless aware that progress in vector design and delivery strategies will be needed before gene therapy can become a realistic therapeutical strategy in humans.