The rising trend in the minimum entrance age for kindergarten in the US has been motivated by findings from cross-sectional studies that older entrants have more favorable school outcomes compared to younger entrants. However, these studies fail to account for endogeneity in entrance age resulting from parental choice, leading to biased estimates of the entrance age effect. This paper uses exogenous variation in birth dates and kindergarten entrance age policies to generate instrumental variable estimates of the effect of delaying kindergarten entrance on children's academic achievement. Both initial level differences and subsequent growth in test scores are examined. I find that entering kindergarten a year older significantly boosts test scores at kindergarten entry. More importantly, entering older implies a steeper test score trajectory during the first 2 years in school. Results also suggest that the benefits from delaying kindergarten entrance tend to be significantly larger for at-risk children.