This study analyzes whether teachers' evaluations of children's academic skills mediate the effect of cultural capital on educational performance. We use ECLS-K data with repeated measures on children in 1st, 3rd and 5th grade and apply fixed effects models, that control for unobserved time invariant child and family characteristics. The analysis shows that teacher evaluations have positive effect on educational performance, but that cultural capital has no effect on teacher evaluations. Instead cultural capital has a direct positive effect on educational performance.