In response to changing climate, river discharge, in particular discharge extremes, may change over the forthcoming decennia. Several hydrological models have become available to analyse the effect of climate change on river discharge in the Netherlands. A water balance model on a monthly basis has been established for the entire Rhine and Meuse basins, and a detailed hydrological model with a daily time step was developed for the river Vecht, a small lowland river in the Rhine basin. Though the current water balance models give a good indication of hydrologic changes in large river basins, their temporal resolution is insufficient to estimate changes in peak flows under changed climate conditions in a direct way. Extending detailed (daily) hydrological models to large river basins demands a huge effort for data collection and modelling, and it is unlikely that such models will be available for targe river basins before the beginning of the next century. To bridge the gap between the coarse water balance models and the required daily time-scale in due time, an alternative approach is attempted using statistical methods for downscaling aggregated discharge series calculated by the coarse models to the required time step. In this paper the effect of different climate scenarios on the discharge regime of the rivers Rhine, Meuse, and Vecht is demonstrated, and estimates for changes in peak flows are given on the basis of the statistical downscaling.