Despite the prevalence of charter schools in the United States, few experimental studies evaluate the impact of charter school attendance on students in the early elementary grades. Using data from a randomized lottery in which kindergarten students and their parents applied to two oversubscribed and well-established charter schools in Denver, Colorado, we estimate the effect of admission to and attendance at these schools on third grade reading and mathematics test scores. We find no evidence of admission or attendance having an impact on third grade achievement but find suggestive evidence of differential effects favoring non-White students. These findings speak to the substantial variation of charter school impacts and raise questions about what families may want from schools aside from or in addition to achievement test impacts.