This study estimates the economic losses attributable to a nonindigenous forest insect, the hemlock woolly adelgid (Adelges tsuga), using cross-sectional and difference-in-difference hedonic price models. The data span a decade of residential property value transactions in West Milford, New Jersey. Hemlock health in naturally regenerated hemlock stands was measured biannually over this period using remote sensing data and the image differencing technique. These data were linked with spatially referenced land use and land cover data, measured twice during the decade, and the locations and housing characteristics associated with parcels sold. Spatial dependence in the regression models was addressed using spatial error and fixed-effects panel data models. The empirical results demonstrated that hemlock decline consistently caused statistically significant reductions in property values both for parcels containing hemlock resources as well as for neighboring nonhemlock parcels. We conclude that failure to account for spatial spillover effects would downwardly bias estimates of economic losses and total economic losses on properties sold during the study period ranged from $0.64 million in the parcel level cross-section model to $2.2 million in the 0.5-km neighborhood difference-in-difference model.