In this paper, we analyze the optimal spatial pattern of open space and residential development in an urban model that includes provision of both local and global public goods. In our model, households choose where to live based on land prices, proximity to employment, and amenity values that include access to open space (local public good). Open space also provides habitat for biodiversity (global public good). We applied the model in the Twin Cities Metropolitan Area and include endogenous land prices, land taxes that finance the purchase of open space, heterogeneous land quality, multiple employment locations, and pre-existing spatial features, such as institutional and environmental amenities. Based on this application we develop an efficiency frontier that shows tradeoffs between maximum welfare of households, which includes provision of local public goods, and provision of habitat for biodiversity, which is assumed to not affect household welfare. We show there is the potential for a large increase in biodiversity conservation with only modest reductions in welfare when starting from a spatial pattern of development that maximizes household welfare. Biodiversity conservation can be improved by changing the spatial configuration of open space towards higher quality habitat and aggregating protected areas to increase contiguity.