This article presents a comparative static framework for predicting the water quality outcomes of water trade. The focus is on a comprehensive treatment of water quality processes. Previous work has assumed that reductions in agriculturally induced water quality externalities are an increasing function of irrigation application rates. The comparative static framework used here allows for the possibility that water transfer can result in both positive and negative water quality externalities as the result of dilution, even when the rate of loading decreases. We apply our model to the Owyhee aquifer of eastern Oregon, an area where nitrate concentrations exceeded the EPA standard of 10 ppm in over 30% of area test wells in 1991. In conclusion, we describe conditions when water trade is most likely to generate positive versus negative water quality externalities. We also draw policy conclusions about the kinds of institutional rules best suited to balance trade-offs between gains to trade, water quality externalities, and transactions costs. ¿ 1999 American Geophysical Union