This research work focuses on the development of an innovative method for enhancing the predictive capability of macro-scale rainfall–runoff models by means of a geostatistical apporach. In our method, one can get enhanced streamflow simulations without any further model calibration. Indeed, this method is neither computational nor data-intensive and is implemented only using observed streamflow data and a GIS vector layer with catchment boundaries. Assessments are performed in the Tyrol region.