A population-based study of pairs of relatives (siblings, second- and third-degree) concordant for multiple sclerosis did not show an excess of like-sexed pairs. In addition, data on human lymphocyte antigen typing for sibling pairs concordant for multiple sclerosis did not find an increase in haplotype sharing for like-sexed pairs. These data do not support the notion that sex-related factors influence multiple-sclerosis susceptibility in families.