AbstractHydrological model testing for spatial and temporal robustness is an important part of the simulation procedure but which faces particular difficulties when a model is intended to simulate surface run-off and other hydrological variables under hypothetical environmental conditions different from the historical ones. These difficulties are dealt with in this paper by constructing a water balance model intended to be used under hypothetical changes of climatic parameters and employing a variation of the differential split sample test. The model is applied to four Greek drainage basins of varying similarity in terms of their specific morphoclimatic characteristics. Results show that the predictive efficiency of the model is not affected by variations in the key climatic parameters, average air temperature and average precipitation, and that the model can be adapted to different basins with coherent adjustments of its calibration parameters.