For this work the Geo-referenced Regional Exposure Assessment Tool for European Rivers (GREAT-ER) was developed further to support river basin management and the implementation process within the EU Water Framework Directive (WFD). This was achieved through predicting spatially resolved pharmaceutical exposure concentrations in whole watersheds. A major focus of this thesis has been placed on modeling challenging watersheds, whereby challenging can refer to hydrological conditions in a watershed as well as to specific emission patterns that occur within the watersheds. The adapted methodology improves the prediction accuracy in such watersheds with GREAT-ER with respect to pharmaceutical exposures, but can also lead to improved results in other application areas. The possibilities of the latest model version are demonstrated by the extensive inclusion of local and regional conditions. In watersheds with highly variable and seasonally changing hydrological situations, GREAT-ER has been applied satisfactorily for the first time, and additionally, the developed approach can be transferred to equivalent watersheds worldwide. Comparison with monitoring data confirms that some of the adjustments have resulted in significantly improved model predictions, especially when hydrological and local conditions are specifically addressed. For example, explicit consideration of local drug emissions from hospitals or private medical practices (e.g., for x-ray contrast agents) can improve predictions at the local scale without compromising regional exposure estimates. Pharmaceuticals that have low concentrations and are barely detectable with established analytical methods can be evaluated with model simulations. In addition, current management strategies implemented under the WFD has been replicated and evaluated. These management scenarios simulated with the model allow an a priori evaluation of risk reduction measures. In combination with targeted monitoring approaches, it was shown that the GREAT-ER model can serve as a valuable tool for exposure and risk assessment of pharmaceuticals even in challenging watersheds. This and the useful combination of targeted monitoring and GREAT-ER simulations and the ability of the modeling approach to predict the expected range of spatial surface water concentrations is demonstrated by three selected journal articles.