Evaluating the environmental and economic impacts of agricultural policies is not a simple task. A systematic approach to policy analysis would include investigating the effect of factors such as tillage practices, crop rotations, type and amount of chemical used, weather, topography, and other soil attributes, on observables such as amount of soil eroded or chemical leached into the groundwater. For comparison purposes, the effects of those factors on the response variable would have to be assessed under alternative policy scenarios. Because the number of factor levels is alarmingly high in most problems, and because policies to be evaluated are often not in place at the time of the study, practitioners have resorted to simulation experiments to generate data. In this paper, we discuss the problem of designing computer simulation experiments, and propose an approach that is based on subsampling the 1992 National Resources Inventory (NRI) points. We apply the procedure to the problem of assessing soil erosion under different policy scenarios.