An export coefficient model of diffuse pollution at large scales has been developed as part of a larger study investigating and mapping the present and future patterns of the quantity and quality of European water resources. The model was developed to provide reasonable estimates across the whole of Europe based on readily accessible datasets, and that would be amenable to application within a gridded model of water quality loadings to surface waters. The export coefficient models for biochemical oxygen demand (BOD), total nitrogen (TN) and total phosphors (TP) were fitted to data from European Union European Environment Agency databases of 79–106 selected river basins around Europe depending on the variable in question. Estimated export coefficients were on a reasonable level with estimates made by other methods within Europe. Firstly, it was shown that runoff, number of livestock and point load were common factors for BOD, TP and TN loads with runoff as the most important factor. Secondly, cropland area also contributed to diffuse TN load. Thirdly, average slope steepness and runoff, as a combined factor, had a negative effect on diffuse TP load. Fourthly, lake area reduced diffuse loads because of retention mechanisms in lakes. A larger set of data with higher spatial and temporal resolution and partitioning of the data based on, e.g. climate or spatial patterns would further improve the precision of the export coefficient estimates. When applied at the catchment scale, the parameters should be updated with local data. In the long run, adaptive modelling should be applied in order to fit the model to continuously accumulating data and experience, although the study showed that the export coefficients derived were stable over time. It is also advisable to integrate the administrative monitoring, modelling and management of river basins more closely to improve data availability, model predictions and cost efficiency of management measures and policies.