The hedonic literature has established that public water bodies provide external benefits that are reflected in the value of nearby residential real estate. The literature has employed several approaches to quantify these nonmarket services. With a residential hedonic model, this paper tests whether model specification affects resource valuation using an actively managed reservoir in Indiana and a passively managed lake in Connecticut. The results indicate that valuation is quite sensitive to model specification,and that omitting either the waterview or waterfront variables from the hedonic function likely results in a misspecified model. The findings from this study are important for researchers and public agencies charged with managing water resources to bear in mind as the external benefits from existing or man-made lakes anr reservoirs are estimated. Therefore, while it requires considerably more effort to determine which properties are in waterfront locations and which properties have a view, the potential misspecification of distance-only models likely justifies these extra research costs. Further, the findings in this analysis callinto question results from distance-only models in the literature.