This article presents a simple long-run multijurisdictional neoclassical model that is used to simulate the incidence of a residential property tax in a semi-open metropolitan area. Housing is produced using capital and land, and households consume housing and a composite good Capital and the composite are in perfectly elastic supply. In the model, a fixed number of households sort themselves across 25 jurisdictions. First, a uniform tax is imposed in all jurisdictions. Next, the tax is raised in one single jurisdiction. The excise effects of raising the tax turn out to be surprisingly large. In the long run, a 25% increase in the tax rate only generates 6.6% more revenue for the taxing jurisdiction. Explicit excess burdens are calculated using the indirect utility approach. © 1991, Sage Publications. All rights reserved.