Nitrogen inputs into surface waters from diffuse sources are still unduly high and the assessment of mitigation measures is associated with large uncertainties. The objective of this paper is to investigate selected agricultural management scenarios on nitrogen loads and to assess the impact of differing catchment characteristics in central Germany. A new modelling approach, which simulates spatially distributed N-transport and transformation processes in soil and groundwater, was applied to three meso scale catchments with strongly deviating climate, soil and topography conditions. The approach uses the integrated modelling framework JAMS to link an agro-ecosystem, a rainfall-runoff and a groundwater nitrogen transport model. Different agricultural management measures with deviating levels of acceptance were analysed in the three study catchments. N-leaching rates in all three catchments varied with soil type, the lowest leaching rates being obtained for loess soil catchment (18.5 kg nitrate N ha(-1) yr(-1)) and the highest for the sandy soils catchment (41.2 kg nitrate N ha(-1) yr(-1)). The simulated baseflow nitrogen concentrations varied between the catchments from 1 to 6 mg N l(-1), reflecting the nitrogen reduction capacity of the subsurfaces. The management scenarios showed that the highest N leaching reduction could be achieved by good site-adapted agricultural management options. Nitrogen retention in the subsurface did not alter the ranking of the management scenarios calculated as losses from the soil zone. The reduction effect depended strongly on site specific conditions, especially climate, soil variety and the regional formation of the crop rotations.