ABSTRACT Land price differentials have long been used as a proxy for the value of environmental improvements in cost/benefit analysis. Both the empirical and theoretical literatures have largely ignored two important facts, however: Taxes financing local improvements are often distortionary, and amenities which influence property values in turn impact the fiscal budget, and hence the tax rate and final economic burden. Put another way, the economic cost of an improvement is endogenous to both the amenity level and the revenue structure. Extending the story in this direction for a system of open or closed spatial cities, the paper finds land rent measures to be a biased measure of the willingness to pay for amenities financed by either head taxes (benefit taxes), property taxes (excise taxes), or highway tolls (user fees). These results are used to correct the conventional specification of empirical property value regression models, which traditionally account for neither tax revenue effects nor the excess burden of distortionary taxation.