Early studies of . the productivity of pesticide expenditures determined a very high marginal product. The marginal product of control will depend upon the level of the particular pesticide, the host crop, initial pest infestation, other state variables, and the functional form. The excessively high marginal product of generic pesticide expenditures estimated in some earlier studies are not evident in our analysis of crop and herbicide specific experimental data. Field-level data for winter wheat in south eastern Washington indicate increasing returns to herbicide application are unlikely to occur in typical field application.