Policymakers have made several attempts to introduce local and national policies to reduce CO2 emissions and stimulate the consumer adoption of alternative fuel vehicles (ethanol/E85 cars). The purpose of this paper is to analyze how a local policy measure impacts the composition of the car fleet over time. More specifically, we take advantage of the natural experiment setting caused by the introduction of the Stockholm congestion tax (2006) to analyze how the tax affected purchases of ethanol cars that were exempted from the tax. To estimate effects, we employ a Difference-in-differences methodology. By using a comprehensive database of the car fleet and car owners, sociodemographic and geographic factors are analyzed, which is unique in the existing literature. Our results suggest that the congestion tax had a significant impact on ethanol car purchases although the effect fades away over time. Furthermore, there is a positive relationship between the level of education and ethanol car purchases. Previous adoption of an ethanol car is found to be the strongest predictor of ethanol car purchases. Finally, data indicate that Stockholmers substantially increased purchases of ethanol cars half a year before the introduction of the congestion tax, which we refer to as an anticipation effect. (C) 2014 The Authors. Published by Elsevier Ltd.