While there is now much evidence in the literature that assignment to ethnically-congruent teachers results in better student outcomes like achievement and teachers' evaluations of their behavior for Black and White students, findings appear to be noticeably mixed for Hispanic students. This paper shows that a potential reason for the mixed findings for Hispanic students lies in the fact that previous studies have not adequately accounted for the cultural background of students and teachers. Unlike existing studies, which define matches to occur when a student and teacher report having the same race, I define matches to occur only if the student and teacher report having both the same race and native language. The rationale is that race and native language together provide a more complete picture of ethnic identity compared to only race. Employing a student fixed effects strategy, and comparing two different teachers' evaluations of the same student, I find that Hispanic students receive more favorable evaluations from Hispanic teachers who share the same native language than Hispanic teachers who speak a different native language or non-Hispanic teachers. This suggests that more coherent findings may emerge if researchers additionally consider native language in defining ethnic matches.