Background: The epidemiology of multiple sclerosis (MS) includes a consideration of geography and population ethnicity. To determine whether there is any association between ethnicity and risk of MS in a multiethnic, population-based case–control study. Methods: We conducted a population-based case–control of 547 incident MS cases and 1057 healthy controls between August 2013 and February 2015, Tehran, a multi-ethnic city. The patients were identified and enrolled through the Iranian MS Society. Case status was confirmed by a panel of MS specialists beside of 2010 McDonald criteria. Controls were selected through random digit dialing. A logistic regression model was applied to estimate the odds ratios (95%CI) adjusted for age, sex, tobacco smoking and socioeconomic status. Results: The risk of MS for Kurd, Turk and Fars ethnicities was apparently smaller compared to one specific ethnic background (Lor); OR 0.36 (0.15–0.86) for Kurd, OR 0.42 (0.24–0.74) for Turk and OR 0.53 (0.31–0.89) for Fars. Heterogeneity in parental ethnicity of common ethnic groups was significantly associated with increased risk of MS OR 1.61 (1.13–2.29). All associations remained after adjustment for relevant confounders. Conclusions: MS was unequally distributed in the ethnic groups. Moreover, heterogeneity in parental ethnicity seems to be a risk factor for MS.