Air pollution in urban centers has become a key public concern around the world. Apart from its adverse health effects, air pollution could impact less visible outcomes like cognitive performance. Standardized tests are a fixture of all education settings that are susceptible to pollution shocks because they require high cognitive function. Given that test scores are widely used as signals by parents, employers, and education institutions, pollution shocks could render tests unfair and unreliable. In this paper, we combine pollution data with test scores from a large testing institution in Iran between 2012 and 2017 to quantify the impact of short-term exposure to air pollution on test performance. We use visibility as an instrument for exposure to PM10 and PM2.5 on the exam day in a student fixed effects specification to study the impact of PM. We find that a one standard deviation increase in PM10 and PM2.5 is respectively associated with 0.047 and 0.029 of a standard deviation decrease in test scores. The effect is larger for males and analytical subjects. The results are robust to the inclusion of other pollutants and several specification checks.