In summary, by comparing the "before" and "after" predicted hydrologic outputs from a coupled stochastic rainfall model and a deterministic method for estimating runoff, a procedure for comparing the hydrologic effects of land use changes on ungauged watersheds has been demonstrated. The procedure is limited primarily by the method of relating rainfall and runoff. However, local data can be used to estimate watershed parameters and thus remove some of the variability of the procedure. While the rainfall model has been validated for only a few areas, the authors feel that this type of model will reduce additional uncertainties in the current practice of implying that rainfall and runoff have similar frequency distributions. Another advantage of the proposed procedure is the flexibility of the previously developed rainfall model. Elevation effects can be taken into account, the distribution function of mean area rainfall can be utilized, the pdf of runoff can be obtained and used in economic evaluations and finally the procedure can generate simultaneous synthetic sets of rainfall, runoff and soil moisture for more detailed hydrologic investigations.