Though there is clinical evidence linking pollution induced inflammatory factors and major depression and suicide, no definitive study of risk in the community exists. In this study, we provide the first population-based estimates of the relationship between air pollution and suicide in the United States. Using detailed cause of death data from all death certificates in the U.S. between 2003 and 2010, we estimate the relationship between daily variation in air quality measured using NASA satellite data, and suicide rates. Using wind direction as an instrument for reducing potentially endogeneity and measurement error in daily pollution exposure, we find that a 1 μg/m3 increase in daily PM2.5 is associated with a 0.49 percent increase in daily suicides (a 19.3 percent increase). We also estimate the impact of days with high air pollution on contemporaneous suicide rates compared to other days in the same state-month, month-year, day of the week and county with lower air pollution, conditional on the same weather and total population. Estimates using 2SLS are larger and more robust, suggesting a bias towards zero arising from measurement error. Event study estimates further illustrate that contemporaneous pollution exposure matters more than exposure to pollution in previous weeks.