Studies of myelin basic protein (MBP)-specific T lymphocytes have been extraordinarily informative for several distinct aspects of (neuro) immunology. They indicated for the first time that myelin directed central nervous system (CNS) inflammation is mediated by specific autoaggressive T cells. This provided a conceptual basis for understanding the pathogenesis of anti-myelin autoimmunity in general. We assume at present that human multiple sclerosis (MS) is also caused by myelin-specific autoaggressive T cells, and that these T cells create the inflammatory changes typical for the generation of a mature demyelinating plaque. Large scale demyelination seems, however, not to be caused by purely cellular mechanisms. Second, studies using MBP-specific T-cell lines have critically contributed to elucidating the immune status of the CNS. They showed that the CNS is not completely secluded from the circulating immune cells. Activated, though not resting lymphocytes can pass through the endothelial blood-brain barrier and migrate through the brain tissues. There, in the CNS, some glia components, microglia and astrocytes can be induced by T-cell cytokines to act as efficient antigen-presenting cells. Thus, the CNS is subject to immune surveillance, though to a specialized, adapted variant thereof. Finally, studies using MBP-specific T cells established for the first time the presence of potentially autoaggressive T cells within the normal immune repertoire. This implies counterregulatory regulation required to keep the autoreactive T cells in a safe state of rest. Most recently, the analysis of the intrathymic anti-MBP T-cell repertoire may change our views on the generation of myelin-specific self tolerance. There is now the possibility that this is by no means based on ‘ignorance’ of a sequestered brain autoantigen, but also depends on intrathymic selection.