Narcolepsy is a neurological disorder known to be associated with human leukocyte antigen (HLA)-DQB1*0602 in humans. In a canine model, the disorder is also genetically linked to a gene of high homology with the human mu-switch-like immunoglobulin (Ig) gene (current LOD score 13.6 at 0% recombination). Since association with HLA or other immune function polymorphic genes (T cell receptor of Ig, mainly) is a hallmark of most autoimmune diseases, it is proposed that autoimmunity may also play a role in the development of narcolepsy. Arguments for and against this hypothesis are reviewed. It is shown that both on the basis of the most recent molecular studies, and because of some of its clinical features, narcolepsy may be an autoimmune disorder. However, neither systemic nor central nervous system (CNS) evidence of any autoimmune abnormality have ever been found. To reconcile this discrepancy, it is suggested that the pathological immune process involved in narcolepsy could be difficult to detect because it is restricted to a very small region of the brain or targets a low abundance neuroeffector. Alternatively, it is possible that a more fundamental relationship is involved between sleep generation and immune regulation. The pathophysiology of narcolepsy may then involve new CNS-immune mechanisms that may shed new light on the sleep process itself.