Cost-effective targeting of conservation activities has only recently been addressed by economists. Most work to date has focused on finding the best locations to set aside land for the protection of biodiversity. An economic approach to the problem, where biodiversity reserve networks are delineated to maximize the number of species protected subject to a budget constraint, has been shown to be much more cost-effective than the standard approach, where reserve networks are delineated subject to an area constraint, ignoring differences in costs across sites. This paper is among the first to use spatially explicit models of production functions for ecosystem services in an optimization framework for prioritizing sites for wetlands restoration. Tradeoffs between two classes of environmental benefits from wetlands restoration, habitat, and water quality were assessed in the Central Valley of California. Habitat benefits were estimated by a count regression model that relates breeding mallard abundances to the configuration of land use types in the study area, and water quality benefits were estimated by a spatially distributed model of nonpoint source pollution and nutrient attenuation in wetlands. Two decision scenarios were analyzed. In the first scenario the optimal configuration of restoration activity was determined for a small watershed, and in the second scenario sites were selected from those offered for enrollment in an easement program throughout the valley. The results reveal the potential for gains in effectiveness from spatial targeting, and they suggest that there will be substantial tradeoffs between environmental benefits. Maximizing habitat quality in the small watershed yielded a 34% increase in mallard abundance and a 3% decrease in nitrogen loads to the river. In contrast, maximizing water quality resulted in a 25% decrease in nitrogen loads and a 2% increase in mallard abundance. Qualitatively similar results were obtained when sites were selected from a set of offered sites throughout the valley, but the tradeoffs were not as severe. The results also suggest that at traditional funding levels the Wetlands Reserve Program in California could reduce nitrogen loads to rivers by approximately 29,000 kg and increase total mallard abundance in the breeding season by approximately 150 individuals throughout the Central Valley in a given year.