The increasing demand for watershed management tools at local and regional scales makes it necessary to develop regionalization techniques for applications of continuous process-based models. To be useful for decision-makers, models must be applicable to all parts of the world with minimum calibration and must consider spatial variability in modeling hydrological and soil erosion processes. We develop regionalization methods for the point model SIMULAT and the hillslope model of the Water Erosion Prediction Project (WEPP). These methods provide model simulations that include spatial variability in detail and are able to reduce time consuming simulation runs by using hydro-pedotopes or hillslope units to represent the hydrological or erosional behavior of a watershed or larger region. Each method is designed to use geo-referenced commonly available data in a Geographical Information System (GIS) by aggregation and disaggregation procedures with a minimum of quality lost in the spatial and temporal distribution of the original input data. The results show that these regionalization methods produce detailed spatially distributed results on a daily basis either for relative water balances components of hydro-pedotopes or soil detachment rates for cells along flow paths on hillslopes within a small watershed that match well with runoff and sediment measurements in both study areas. The evaluation of different quality of topographical model input on regionalization results demonstrates the importance of different data quality and scales on model assessment results and decision-making.