lthough organic cropping systems are promoted for their environmental benefits, little is known about their long-term impact on nitrogen (N) fate in the soil–plant-atmosphere system. In this paper, we analyze two long-term experiments: DOK in Switzerland (39-yr) and Foulum organic in Denmark (19-yr). Four treatments were considered in each experiment: two conventional treatments with (CONFYM) or without manure (CONMIN), organic with manure (BIOORG) and unfertilized treatment (NOFERT) at DOK; conventional (CGL-CC+IF) and three organic treatments, one with cover crops only (OGL+CC-M) and two including cover crops and grass-clover with (OGC+CC+M) or without manure (OGC+CC-M), at Foulum. STICS model was used to simulate crop production, N surplus, nitrate leaching, gaseous N losses and changes in soil organic N. It was calibrated in the conventional treatments and tested in organic systems. The crop production, N surplus and soil organic N stocks were satisfactorily predicted. The mean N surplus greatly differed between treatments at DOK, from −58 (NOFERT) to +21 kg N ha−1 yr−1 (CONFYM), but only from −9 (OGL+CC-M) to +21 kg N ha−1 yr−1 (OGC+CC+M) in Foulum. Soil N pools declined continuously in both sites and treatments at a rate varying from −18 to −78 kg N ha−1 yr−1, depending on fertilization and crop rotation. The decline was consistent with the observed N surpluses. Although not all simulations could be tested against field observations and despite of prediction uncertainties, simulations confirm the hypothesis that environmental performances resulting from C and N cycles depend more on specificities of individual than nominal treatments. Significant correlations appeared between long-term N surplus and soil N storage and between total N fertilization and total N gaseous losses. Results showed in both experiments that arable organic systems do not systematically have lower N surplus and N losses than conventional ones, providing opportunity for increasing N use efficiency of these systems. © 2019 Elsevier B.V.