Promoting biological weed regulation by shifting resource availability and use from weed to crop may provide an option for a more sustainable weed management. Light is generally the main resource for which crops and weeds compete in conventional cropping systems. But, with the necessity to reduce mineral nitrogen fertilizer use, better managing crop-weed competition for nitrogen may become crucial. However, it requires better understanding the functioning of heterogeneous canopies in nitrogen-deficient situations. Simulation models are powerful tools to reach this goal. Our objective was to integrate plant-plant competition for nitrogen into the FlorSys model, already simulating competition for light. The final aim was to provide the first mechanistic and 3D individual-based model simulating the effects of cropping system and pedoclimate on weed dynamics, integrating competition for nitrogen. The new formalisms were mostly inspired from pre-existing models and adapted to make them compatible with the individual-based representation of FlorSys. Soil-nitrogen concentration is predicted by the STICS soil submodel linked to FlorSys. Plant nitrogen uptake was simulated by confronting plant nitrogen demand (driven by shoot growth) to plant nitrogen supply (depending on root characteristics, soil-nitrogen availability and the presence of neighboring plants with roots in the same soil zone). Competition for nitrogen occurred when the amount of nitrogen available in a soil voxel (i.e. 3D soil pixel) was lower than the requirements of all the plants with roots in this voxel. A nitrogen stress index allowed to account for the impact of plant nitrogen nutrition on plant photosynthesis, biomass allocation and morphology. To reflect the plant adaptation to the spatial heterogeneity in soil-nitrogen availability, we introduced ‘compensation’. For a given plant, if nitrogen uptake in one soil voxel is insufficient to fulfil plant nitrogen requirements in this voxel, this local nitrogen- deficiency could be compensated by increasing nitrogen uptake in other nitrogen-richer voxels. The new formalisms needed only seven plant parameters which we measured for several crop and weed species. Simulations showed that, despite simplifying hypotheses in formalisms, predictions were consistent with knowledge on canopy functioning and crop-weed interactions. The nitrogen version of FlorSys will be useful to understand the role of nitrogen in crop-weed interactions and to identify sustainable management strategies promoting weed regulation by competition (see Perthame et al., this congress). Due to its process-based representation and genericity (it can simulate diverse crop species), it will also be useful to better understand crop-crop interactions in intercropping.