In the search for the etiology of multiple sclerosis (MS), consideration has been given in turn to infectious agents, to genetic markers, and more recently to a combination of genetic and environmental factors, but after over a century of research, a definite conclusion has not been reached. The hypothesis that an infectious agent is responsible for triggering MS is perhaps one of neurology's most enduring notions. Interest in an infectious etiology has waxed and waned over the last two centuries since Pierre Marie first proposed that MS often starts as an infectious process. The possible role of infectious agents has been suggested by: the different geographic gradients in frequency among Caucasians; changes in prevalence due to migration, and the effect of age at migration; the suggestion of epidemics and clusters of cases in some small communities; and the remarkably low degree of concordance in monozygotic twins. The infectious hypothesis is strongly supported by the different temporal patterns of the disease in different geographic areas. Incidence rates have remained stable in some areas, but have changed over time in other regions. On the other hand, the hypothesis is hampered by the lack of evidence for a specific agent, and the weakness of the results of analytical studies that have tested the association between MS and previous infections. Despite these drawbacks, recent studies of a few select pathogens suggest that viral or bacterial infections or reactivations may trigger clinical exacerbations in relapsing-remitting MS.