A model based on graph theory was developed to efficiently evaluate the impact of the effluent from 742 wastewater treatment plants (WWTPs) on micropollutant loading throughout all river catchments in Switzerland. Model results agree well with measured loads for 12 compounds in river water samples, revealing mean predictive accuracy factors between 0.8 and 3.4. Subsequently, pollutant concentrations were predicted for river sections downstream from 543 WWTPs where hydrological information was available, and compared with recent recommendations for water quality criteria. At base flow conditions, carbamazepine concentrations (parent compound only) are ubiquitously below a water quality criterion of 0.5 microg L(-1). In contrast, the sum of diclofenac and its metabolites is expected to exceed the corresponding water quality criterion of 0.1 microg L(-1) in 224 river sections. If diclofenac cannot be eliminated atthe source, the model suggests a directed upgrade of 173 WWTPs to meet the condition that concentrations are never to exceed this water quality criterion.