This paper explores the influence of spatial mobility on the risk of overeducation of French young workers. We mobilise a survey that follows a cohort of young people who graduated in 2010 until 2013. The effect of residential migration on the probability of being overeducated is estimated through linear probability models. Our analysis deals with self-selection into employment and endogeneity of mobility decision by combining Heckman procedure and instrumental variables method. Estimated results reveal that regional migration decreases the risk of (statistical and subjective) overeducation. We also evidence differentiated effects for migration to Paris and/or according to the educational level.