Background: Stress has been considered a triggering factor for multiple sclerosis (MS) since the description of the disease by Jean-Martin Charcot. Until our times, many published studies have supported that both MS onset and relapse could be predisposed by psychological stress. This review aims to synthesize existing knowledge of the relationship between psychological stress and MS onset and relapse, focusing mainly on the quality of observational studies. Methods: We hand-searched MEDLINE with the terms ‘stress and multiple sclerosis’, using English language restrictions, from January 1980 to November 2010. We included only observational longitudinal studies. The Newcastle-Ottawa scale proposed by the Cochrane Collaboration was used for assessing the quality of the observational studies. Results: Seventeen publications were analyzed, 5 for MS onset (1 cohort and 4 case-control studies) and 12 for MS relapse (9 cohort and 3 case-control studies). We found a marked heterogeneity in stress measurement that mostly targeted the environmental approach to stress. Only 2 publications used radiological criteria for MS relapse. Quality issues were identified mainly for comparability, meaning that studies failed to control adequately for various triggering and psychosocial factors in the stress-MS relationship. Also, selection and blinding problems were identified in most case-control studies. All studies, with only 2 exceptions, resulted in favor of the stress-MS relationship, but due to marked stress measurement heterogeneity, no secure conclusions could be drawn. Conclusions: Future studies should incorporate a multidisciplinary approach to stress measurement and radiological criteria for MS. We further encourage researchers to test the effect of early life stress and stress management techniques on the clinical course of the disease.