With efficiency standards and consumer information, car-related taxes are one of three pillars of the European Commission's strategy to reduce CO2 emissions from passenger cars. A long-standing question concerns the effectiveness of such taxes in determining the car-purchasing behavior of households. Several recent studies suggest that purchases are primarily determined by retail cost rather than by taxes, which typically are incurred over the lifetime of the car. Panel data on new car registrations in Germany, Europe's largest car market, were used to address this issue through an econometric analysis of the impact of fuel costs and circulation taxes on car market shares. By using a nested logit model that explicitly recognizes the segmented structure of the car market, the analysis considers correlation in unobserved shocks in cars belonging to the same market segment. Moreover, given the panel structure of the data, a fixed-effects estimator is used to control for the influence of unobservable, time-invariant automobile attributes that could otherwise induce biases in the estimated coefficients. Contrasting with much of the evidence garnered to date, the results suggest that circulation taxes and fuel costs significantly determine car market shares and hence may serve as effective instruments in influencing the composition of the car fleet and associated CO2 emissions.