Recent literature on the Environmental Kuznets Curve (EKC) has raised questions about the strength of the empirical results, the appropriate model specification, and the validity of the underlying theoretical model. Data limitations and measurement error have also been issues; many empirical studies of air pollution have utilized the GEMS/AIR dataset which extends only to 1993 and is considered by some observers to be unreliable. The current analysis utilizes a newly compiled dataset drawing on spatial data that includes more than 20,000 observations for air quality in Europe and North America. The model specification builds on specific theoretical findings. Results for SO2 and PM10 air pollution exhibit stronger EKC relationships than have generally been found in the prior literature. The significance of these results for further research and for their policy implications are discussed.