The effect of state funded merit aid for college tuition on student performance in high school has been largely ignored by the current literature on merit aid programs, even though one aim of these programs is to provide incentives for increased academic achievement in high school. To test for such effects, we employ a selection model for graduation rates using the National Center for Education Statistics (NCES) Common Core of Data (CCD) on over 8,500 school districts for 1990 and 2000. The results show that graduation rates improved in merit aid states relative to non-merit aid states over this period. Perhaps surprisingly, African American students appear to have benefited from merit aid's effect on graduation rates to the same or a greater extent than other students.