The role of autoantibodies in multiple sclerosis (MS) has been enigmatic since the first description, many decades ago, of intrathecal immunoglobulin production in people with this condition. Some studies have indicated that MS pathology is heterogeneous, with an antibody-associated subtype - characterized by B cells (in varying quantities), antibodies and complement - existing alongside other subtypes with different pathologies. However, subsequent evidence suggested that some cases originally diagnosed as MS with autoantibody-mediated demyelination were more likely to be neuromyelitis optica spectrum disorder or myelin oligodendrocyte glycoprotein antibody-associated disease. These findings raise the important question of whether an autoantibody-mediated MS subtype exists and whether pathogenic MS-associated autoantibodies remain to be identified. Potential roles of autoantibodies in MS could range from specific antibodies defining the disease to a non-disease-specific amplification of cellular immune responses and other pathophysiological processes. In this Perspective, we review studies that have attempted to identify MS-associated autoantibodies and provide our opinions on their possible roles in the pathophysiology of MS.