The positive impact that competition has on performance in most industries has been questioned in the education sector. The difficulty to measure competition, the idea that parents don't rationally choose schools for their children, and that schools do not react to that choice is in the center of the debate. We critically analyze the prevailing methodology in the literature that relates competition and educational performance, and the data used to estimate that impact. We propose a methodology that considers relevant substitutes for each school using various attributes which parents consider when choosing schools, and we apply it to estimate the effect of competition on educational performance in Chile, were more than 90% of the students are covered by a voucher. The evidence supports the hypothesis that competition has a positive, statistically significant, and economically relevant educational impact on private and public schools.