In this study, we provide evidence that parents' beliefs about the value of math, in terms of successful employment, have a positive impact on children's math scores. This result is robust to the reverse causality issue that characterizes the relationship between parental attitude and children's performance. We adopt an identification strategy that relies on two pillars. First, using PISA 2012, we estimate this relationship on a sample of immigrants that includes second-generation students and first-generation students who migrated before starting primary education. Second, we instrument the parental attitude with the country of origin math performance, under the assumption that country of origin math performance affects children's performance only through parents. We find that one additional score point in the origin country performance in math increases student performance by 21 percent of one standard deviation of the student math score. For an indirect transmission mechanism through parents math culture, this can be considered a quite substantial effect. Disentangling the effect of one of the factors that shape the family background, we contribute to the empirical literature on the explanations of individual educational achievements.