For more than ten years, scientists of the Universities of Karlsruhe, Heidelberg and Bayreuth have investigated the water dynamics as well as transport of matter in the Weiherbach catchment, a loess area of about 6.3 km 2 located in the ''Kraichgau'' of Southwest Germany. The investigations had the following scientific objectives: to identify the processes governing water and solute dynamics on different spatial and temporal scales in a small rural loess catchment, to develop a physically based numerical model to cope with all hydrological situations, i.e. simulating the extreme cases of very small runoff and flood runoff events as well as simulating soil water flow and transport, based on the obtained understanding of processes as well as to provide a data set for hydrological research for the intensively used rural loess areas of Central Europe.Modeling approaches for the processes dominating water and mass balance on the plot, hillslope and small catchment scale were developed based on process studies, tested against observations and implemented into the physically based, distributed model CATFLOW. During the observation period rainfall-runoff activity was low. The distribution of soil types was found to be strongly influenced by erosion leading to a typical hillslope soil catena. Preferential flow in earthworm burrows turned out to be crucial for solute transport in the soil on the plot and the hillslope scale, especially for leaching of surface applied pesticides, but also for runoff generation on the catchment scale. However, pesticide as well as phosphate loads in the Weiherbach creek during rainfall-runoff events were strongly determined by losses of sewer pipelines, which drain paved farm courtyards and pesticide loss due to runoff from courtyards. Those entrance paths are difficult to model in physical terms.A simplified, effective model approach for preferential flow based on field and laboratory measurements yielded simulation results in good accordance with short term observations of tracer transport on the plot scale as well as long term observations of tracer transport on the hillslope scale. Assuming that the soil catena of each hillslope may be substituted by the same typical soil catena a long term simulation of catchment scale water dynamics yielded results in acceptable agreement with the observed rainfall-runoff dynamics, soil water dynamics as well as evapo-transpiration. However, a falsification of the developed modeling approach was not possible, because of the uncertainty of the obtained parameter values due to measurement errors and due to the enormous variability of state variables and parameters in the Weiherbach soils.