Early school times fundamentally clash with the late sleep of teenagers. This mismatch results in chronic sleep deprivation, which poses acute and long-term health risks and impairs students' learning. Despite conclusive evidence that delaying school times has immediate benefits for sleep, the long-term effects on sleep are unresolved due to a shortage of longitudinal data. Here, we studied whether a flexible school start system, with the daily choice of an 8AM or 08:50AM-start, allowed secondary school students to improve their sleep and psychological functioning in a longitudinal pre-post design over exactly 1 year. Based on 2 waves, each with 6-9 weeks of daily sleep diary, we found that students maintained their 1-hour-sleep gain on days with later starts, both longitudinally (n=28) and cross-sectionally (n=79). This sleep gain was independent of chronotype and frequency of later starts but differed between genders. Girls were more successful in keeping early sleep onsets despite later sleep offsets, whereas boys delayed their onsets and thus had reduced sleep gains after 1 year. Students also reported psychological benefits (n=93), increased sleep quality and reduced alarm- driven waking on later school days. Despite these benefits on later schooldays, overall sleep duration was not extended in the flexible system. This was likely due to the persistently low uptake of the late- start option. If uptake can be further promoted, the flexible system is an appealing alternative to a fixed delay of school starts owing to possible circadian advantages (speculatively through prevention of phase-delays) and psychological mechanisms (e.g. sense of control).