The objective of this paper is to estimate the effects, both aggregated and disaggregated by trimester, of in utero lead exposure on birth outcomes. These outcomes are not well understood. Our identification strategy exploits National Association for Stock Car Racing’s (NASCAR) voluntary switch from leaded to unleaded racing fuel in 2007, more than a decade after the U.S. had otherwise banned leaded gasoline for on-road use. We examine 147,673 births from the Charlotte-Concord-Gastonia Metropolitan Statistical Area in North Carolina from 2004-2009 and use a quasi-experiment exploiting the NASCAR switch from leaded to unleaded racing fuel in 2007. We use a multivariate regression model, controlling for potentially confounding variables, combined with coarsened exact matching, and estimate a differencein-differences model to obtain the average treatment effect on the treated from the deleading policy on birth outcomes around the Charlotte Motor Speedway. Our results indicate that after deleading, children born to mothers residing within 4,000 meters of the racetrack experienced an average increase in birth weight of 103.9 grams and gestational age by 0.36 weeks. The probability of low birth weight declined by 4.1 percentage points, preterm births by 2.7 percentage points, and small for gestational age by 4.1 percentage points. Exploiting variation in the timing of racing events across trimesters revealed heterogeneous effects for birth weight, gestational age, and preterm birth, with first trimester exposures having the largest impact. The implication of our results is that reducing exposure to even short-term lead emissions can significantly improve infant health outcomes.