We examine the impact of US colleges and universities switching from an academic quarter calendar to a semester calendar on student outcomes. Using panel data on the near universe of four-year nonprofit institutions and leveraging quasi-experimental variation in calendars across institutions and years, we show that switching from quarters to semesters negatively impacts on-time graduation rates. Event study analyses show that these negative effects persist well beyond the transition. Using detailed administrative transcript data from one large state system, we replicate this analysis at the student-level and investigate several possible mechanisms. We find shifting to a semester: (1) lowers first-year grades; (2) decreases the probability of enrolling in a full course load; and (3) delays the timing of major choice. By linking transcript data with the Unemployment Insurance system, we find minimal evidence that a semester calendar leads to increases in summer internship-type employment.