Early school times clash with the late sleep of adolescents, leading to wide-spread sleep restriction in students. Evidence suggests that delaying school starts is beneficial for sleep and recent studies investigated whether this also translates into improved academic achievement. We thus conducted a systematic review of the literature on school start times, grades and test scores in middle and high-school students. We reviewed 21 studies following the PRISMA guidelines and assessed the evidence quality using a pre-defined risk of bias tool. Nine studies reported no association of later starts with achievement, while the remaining reported mixed (5), positive (5), negative (1) or unclear (1) results. Considering the heterogeneity in academic outcomes, study types, amount of delay and exposure, and the substantial risk of bias, a meta-analysis was not warranted - instead we provide grouped reviews and discussion. Overall, no generalisable improvements in achievement with later starts emerge beyond the level of single studies. This does not necessarily preclude improvements in students’ learning but highlights shortcomings of the literature and the challenges of using grades and test scores to operationalise academic achievement. Given other previously reported positive outcomes, our results suggest that schools could start later while achievement is likely maintained.