This paper presents a spatially explicit analysis of the contribution of urban trees and parks to residential property values. We estimated the effects of structural, neighbourhood, and environmental variables, including tree cover, on sale price of single-family homes in Perth using a generalized spatial two-stage least-squares model. The spatial model results indicate that, among other structural and neighbourhood variables, the proportion of tree cover on street verges (public space) and the extent and proximity of neighbourhood parks attract significant price premiums in the Perth housing market. However, we failed to find any evidence of impact of the tree cover on property (private space) on its sale price. Further, we find that the parameter estimate for street tree cover obtained from spatial hedonic model is half the size of estimate obtained from ordinary hedonic model, thus indicating importance of spatially explicit hedonic model on implicit price estimation. Based on the spatial hedonic model, it is estimated that 10% increase in tree cover on street verge above the median cover of 19.66% increases the median house price (AU$ 765,000) by about AU$ 3,250. Our findings have implications on managing and developing urban forest cover along the streets in Perth for both private and public benefits.