Nitrogen is essential to improving agricultural production systems, explaining why the contamination of groundwater by this nutrient is widespread. The aim of this paper was to describe data collected over 9 years on estimated leaching levels using a simple computation procedure and measurements of soil water content and water balance, nitrate concentrations in drainage water and meteorological data, and to assess the impacts of the duration of the grassland phase and the level of nitrogen fertilization on grassland on the drainage water quality. The study was carried out at a site of the Long-term Environmental Research Observation and Experimentation facility (SOERE) for Environmental Research-Agro-ecosystems, Biogeochemical Cycles and Biodiversity, run by the INRA experimental unit of Lusignan. The experimental treatments were sequences of maize, wheat and barley with different grassland rotational periods (a pure arable crop rotation; three or six years of grassland receiving high-level nitrogen applications; six years of grassland with a low N application rate and long-term grassland with nitrogen application). The study covered the period from April 2005 to June 2013, during which most drainage occurred in the autumn and early winter. Treatments with the longest duration of grassland exhibited less drainage than those containing a higher proportion of arable crops. The average nitrate concentration was 52.7 +/- 38.63 mg NO3 L-1 under a pure crop rotation, compared to 14.9 +/- 14.76 mgNO(3) L-1 under a petmanent grassland treatment. There were significant differences (P<0.0001) in cumulative nitrogen leaching between the different cropping systems, ranging from 9 to 37 kgN ha(-1) year(-1). The introduction of mowed grassland sequences into this arable crop rotation caused a marked reduction in the nitrate levels in groundwater, and the greater the proportion of grassland within the rotation the more markedly was the NO3- concentration reduced, whatever the level of N fertilization during the grassland sequence.