To study and predict environmental impacts of land use changes on water quality we conduct different types of lysimeter experiments. All of them are linked to representative experimental catchment areas in the field. This allows the verification and extrapolation of lysimeter results. The objective of this paper is to discuss a strategy for using and scaling up of lysimeter results to a field and catchment scale. It will be shown that the N-loss determined with lysimeters falls within the variation of an N-balance based model calculations also to ground water recharge rates calculated with models commonly used in hydrology. Extrapolation of lysimeter data to a catchment with similar soils provides a reliable basis for estimating the N-leaching caused by a change in agricultural land use. On the basis of the N-loss from the soil and the N-load of the stream, the calculations show that an increase in the proportion of one year rotation follow from 10 to 25 per cent results in nearly a 10 per cent increase in the N-load of the stream. From this it can be concluded that set aside in the form of rotation fallow for only one year does not help to protect water quality.