This study uses exogenous variation in sibling sex composition to estimate the causal effect of sibship size on boys' probabilities of private school attendance and grade retention. Using the 1990 U.S. Census, we find that for second-born boys, increased sibship size reduces the likelihood of private school attendance by six percentage points and increases grade retention by almost one percentage point. Sibship size has no effect for first-born boys. Instrumental variable estimates are largely consistent across racial groups, although the standard errors are larger for nonwhites as they have smaller sample sizes and this renders them insignificant at traditional alpha levels. (Contains 6 tables and 7 footnotes.)