Car exhaust is a major source of air pollution, but little is known about its impacts on population health due to socioeconomic selection, measurement error, and avoidance behaviors. We exploit the dispersion of emissions-cheating diesel cars-which secretly polluted up to 150 times as much as gasoline cars-across the United States from 2008-2015 as a unique opportunity to overcome these empirical challenges and measure the health impacts of car pollution. Using the universe of vehicle registrations, we demonstrate that a 10 percent cheating-induced increase in car exhaust increases rates of low birth weight and acute asthma attacks among children by 1.9 and 8.0 percent, respectively. These health impacts occur at all pollution levels and across the entire socioeconomic spectrum. * Diane Alexander (dalexand@frbchi.org) is an economist at the Federal Reserve Bank of Chicago. Hannes Schwandt (schwandt@northwestern.edu) is an assistant professor of economics at the School for Education and Social Policy at Northwestern University. We are thankful to 1 threats to causal inference, including socio-economic selection, avoidance behavior, and measurement error. In this paper, we exploit a unique natural experiment that overcomes these empirical challenges.In 2008, a new generation of supposedly clean diesel passenger cars was introduced to the U.S. market. 3 These new diesel cars were marketed to environmentally conscious consumers, with advertising emphasizing the power and mileage typical for diesel engines in combination with unprecedented low emissions levels. Clean diesel cars won the Green Car of the Year Award in 2009 and 2010, and quickly gained market share. By 2015, over 600,000 cars with clean diesel technology were sold in the United States. In the fall of 2015, however, it was discovered that these cars covertly activated equipment during emissions tests that reduced emissions below official thresholds, and then reversed course after testing. In street use, a single "clean diesel" car could pollute as much nitrogen oxide (NO X ; a precursor to fine particulate matter and ground-level ozone) as 150 equivalent gasoline cars. 4 Hereafter, we refer to cars with "clean diesel" technology as cheating diesel cars.The dispersion of these cheating diesel cars across the United States gives us a unique opportunity to measure the causal effect of car pollution on infant and child health. This natural experiment provides several unique features. First, it is typically difficult to infer causal effects from observed correlations of health and car pollution, as wealthier individuals tend to sort into less-polluted areas and drive newer, less-polluting cars. The fast roll-out of cheating diesel cars provides us with plausibly exogenous variation in car pollution exposure across the entire socioeconomic spectrum of the United States. Second, it is well established that people avoid known pollution, which can mute estimated impacts of air pollution on health (Neidell, 2009). Moderate pollution increases stemming from cheatin...