Research on the effect of language skills on earnings is complicated by the endogeneity of language skills. This study exploits the phenomenon that younger children learn languages more easily than older children to construct an instrumental variable for language proficiency. We find a significant positive effect of English proficiency on wages among adults who immigrated to the United States as children. Much of this effect appears to be mediated through education. Differences between non-English-speaking origin countries and English-speaking ones that might make immigrants from the latter a poor control group for nonlanguage age-at-arrival effects do not appear to drive these findings.