Background and aims: In legume- based intercrops, nitrogen is supposed to be transferred from the legume to the companion crop. However, nitrogen transfer in annual intercrops remains difficult to assess and poorly documented. We aimed to quantify bi-directional transfer between durum wheat and field pea using 15N labeling methods. Method: In a first greenhouse experiment, plant species were grown alone or intercropped. Bidirectional transfer between crops was assessed at three growth stages. In a second experiment, we compared the amounts of nitrogen transferred between species with either intermingled roots, or root separated by a fine mesh, with two levels of soil N. Results: In both experiments, the amount of N transferred from the legume to the cereal was small and did not differ from that transferred from the cereal to the legume. T ransfer between species was significantly higher when roots were intermingled than when they were separated by a mesh. It was smaller in intercrops supplied with nitrate than in unfertilized pots. Conclusion: Since nitrogen net flux from the legume to the cereal was null, N transfer in legume based intercrops may have no effect on the cereal yield.