This paper uses public schools data to examine the effects of age on student academic achievement in light of recent trends to delay children's entry into kindergarten. To avoid problems of endogeneity, students' quarters of birth is employed as an instrument for age at entry. In particular, the effects of students' quarters of birth on math and reading standardized test scores and grade retention at the elementary school level are estimated. Evidence of benefits to being among the oldest in one's age-grade cohort is found. Interestingly, over time, the youngest students begin to perform approximately on par with the oldest students, creating a u-shaped pattern to the data. (Contains 6 tables.)