This paper analyses the impact of the 2003 CAP reform (the so-called Fischler Reform) and its interaction with the Nitrate Directive on the sustainability of selected arable farming systems in a French region (Midi-Pyrénées). The Nitrate Directive is one of the oldest EU environmental programs designed to reduce water pollution by nitrate from agricultural sources, through a set of measures, defined at regional level, and mandatory for farmers of vulnerable zones. This impact analysis is performed through a bio-economic modelling framework coupling the crop model CropSyst and the farm-based model FSSIM developed, within the EU FP6 SEAMLESS project (Van Ittersum et al., 2008). The 2003 CAP reform was compared first to the continuation of Agenda 2000 Regulations and then to a policy scenario combining the CAP reform with the application of the Nitrate Directive. Compared to the continuation of Agenda 2000 Regulations, the implementation of the 2003 CAP reform leads to (i) a decrease of durum wheat area, as the supplement for durum wheat in traditional production zones was reduced and integrated in the single payment scheme, (ii) a slight increase in the land used for irrigated crops, especially for maize grain, considering that 25% of the payments for these crops remain coupled and (iii) an amelioration of farm income due to a better crop allocation. Regarding the environmental results, the 2003 CAP reform induces a decrease of nitrate leaching mostly because of the drop in the level of durum wheat growing under cereal rotations in profit of soft wheat-sunflower rotation which generates less pollution levels. The impact analysis of the policy scenario shows that the potential 3% premium cut is not enough to compel farmers to adopt the Nitrate Directive and to substitute entirely the current activities by the alternative ones based on better N management. The farm income is marginally affected in spite of this premium cut thanks to the implementation of certain alternative activities which are more competitive. The impact on nitrate leaching is not always positive and swings between -6% to +5% depending on farm types. This implies that the partial adoption of better N management is not sufficient to ensure a reduction of leached nitrate. A sensitivity analysis shows that 17% of premium cut is required to enforce all arable farmers in the region to implement this directive.