We study the impact on student achievement of a nationwide teacher pay-for-performance program implemented in Peruvian public secondary schools in 2015. Schools compete in a tournament primarily based on 8th graders' performance in a standardized test, where the principal and teachers of the top 20 percent of schools receive a substantial bonus. We perform a difference-in-differences estimation comparing the internal grades of 8th and 9th graders of the same school, before and after the program. We find a precisely estimated zero effect on student achievement, and we reject impacts greater than 0.017 standard deviations, well below those previously found in the literature. We provide evidence against a series of potential explanations, and argue that this zero effect could be a consequence of teachers' uncertainty about how to improve their students' performance in the standardized test tied to the bonus.