The layout of an urban area can impact air pollution via changes in emissions and their spatial distribution. Here, we explore relationships between air quality and urban form based on cross-sectional observations for 111 U.S. urban areas. We employ stepwise linear regression to quantify how long-term population-weighted outdoor concentrations of ozone, fine particulate matter (PM 2.5), and other criteria pollutants measured by the U.S. Environmental Protection Agency depend on urban form, climate, transportation, city size, income, and region. Aspects of urban form evaluated here include city shape, road density, jobs-housing imbalance, population density, and population centrality. We find that population density is associated with higher population-weighted PM 2.5 concentrations (p < 0.01); population centrality is associated with lower population-weighted ozone and PM 2.5 concentrations (p < 0.01); and transit supply is associated with lower population-weighted PM 2.5 concentrations (p < 0.1). Among pollutants, interquartile range changes in urban form variables are associated with 4%-12% changes in population-weighted concentrations-amounts comparable, for example, to changes in climatic factors. Our empirical findings are consistent with prior modeling research and suggest that urban form could potentially play a modest but important role in achieving (or not achieving) long-term air quality goals.