Recent studies suggest that a history of cigarette smoking is a risk factor for multiple sclerosis (MS). We aimed to test the smoking effect in multiplex families, matching for both environmental and genetic factors. In a matched case-control study, 136 MS patients from 106 multiplex MS families were compared with their 204 healthy siblings as controls. Participants completed self-report questionnaires. Conditional logistic regression was used to analyse smoking and MS risk association while controlling for confounding by age and sex. Smoking history was classified in different variables. Within our survey the smoking history of MS patients and the controls did not differ. The odds of MS were comparable for different smoking levels. However, more intense exposure and women showed higher odds ratios, although non-significant. Association studies in families with relatively high genetic loading are unlikely to be confounded by smoking history.