An empirical investigation of trade-offs between number of children and their scholastic performance confirms that family size directly affects children's achievement. Though parents show no favoritism to first-born children, being early in the birth order implies a distinct advantage, entirely because of the higher probability of being in a small family. recent large changes in family size explain a portion of aggregate test scores declines, but increased divorce rates and market work by mothers have no apparent impact. Finally, teachers are shown to differ enormously, even though performance differences are poorly captured by commonly measured teacher characteristics. The evidence supports a teacher skill interpretation of differences in classroom achievement.