Recent studies of the valuation of real estate have concentrated on the use of hedonic pricing techniques in which the implicit prices of the component characteristics of an asset are inferred from the observed sale price using regression analysis. All of these studies include as explanatory variables one or more locational factors, such as distance to the central business district, as proxies for the effect that location has on the utility of land. In this research, the explicit consideration of the location of real estate in terms of the geographic or Cartesian coordinates (spatial attributes) of observed sales is shown to be a potential substitute for such proxies, either wholly or in part. Such use of spatial attributes could improve the usefulness of the hedonic methodology while at the same time significantly reducing cost and eliminating sources of error.