This article addresses the question of whether youth drinking leads to lower grades for public and private school students. Instrumental variables (IV) regressions are used to investigate differences in the relationship between drinking and grades based on whether students attend a public or private school. The regressions reveal that drinking lowers the probability of earning an "A" average, and raises the probability of earning a "C" or below, for public school students. However, the results are generally not significant for private school students and are sensitive to instrument selection, while empirical testing casts doubt on instrument exogeneity in some instances. (Contains 7 tables.)