This paper uses a unique policy change in Canadaâ€™s most populous province, Ontario, to provide direct evidence on the effect of reducing the length of high school on labour market outcomes for high school graduates. In 1999, the Ontario government eliminated the fifth year of education from its high schools, and mandated a new four-year program. This policy change created two cohorts of students who graduated from high school together with different amounts of education, thus making it possible to identify the effect of one extra year of high school education on earnings. Using restricted survey data, the results demonstrate that students who receive one less year of high school education receive wages that are approximately ten percent lower than their counterparts one year after graduation, and these effects persist two years after graduation. Using birth year to instrument for educational attainment produces estimates that are even higher than the cross-sectional findings, but quite consistent with the existing literature on the return to education. These results are a significant contribution to the literature on the return to education because unlike prior changes to the educational system, this change in schooling laws results in two cohorts entering the labour market simultaneously. As such, business cycle effects do not confound the results.