Multiple sclerosis (MS) is regarded as an autoimmune disease that leads to chronic inflammation in the central nervous system (CNS), and is associated with demyelination and secondary neurodegeneration [1]. This view is mainly based on results obtained using models of experimental autoimmune encephalomyelitis (EAE), the prototypic model of immune-mediated inflammation and tissue injury in the CNS [2]. As direct studies on disease mechanisms in MS patients are difficult or even impossible, most of our current concepts of disease pathogenesis are based on knowledge obtained using these models. There is good evidence that an MS-like inflammatory demyelinating disease is induced in models of EAE, and that active sensitization with brain tissue can directly lead to a disease in humans that, in essential features, reflects MS. Thus, immunization of humans with brain tissue in the course of rabies vaccination or “fresh cell therapy” induced a subacute disease of the nervous system, which pathologically closely resembled MS [2–4]. Preclinical testing of new treatments is also commonly performed in EAE models, and has led to the successful development of a number of current MS therapies. The most impressive example of this is the effective blockade of inflammation and disease exacerbation by natalizumab, a monoclonal antibody that is directed against the adhesion molecule α4 integrin. α4 integrin is instrumental in leukocyte migration into and within the CNS [5–7]. Despite this undisputed success, a number of recent clinical trials of treatments for MS, which used strategies that were highly successful in EAE models, have failed in MS patients. A detailed account of the therapeutic benefits and failures of these studies is provided in a number of excellent reviews [8–11].