Improved agricultural practices are encouraged to reduce nitrate leaching and greenhouse gas emissions. However, the effects of these practices are often studied at annual or rotation scale without considering their long-term impacts. We have evaluated the effects of catch crops (CC), no-till (NT) and reduced nitrogen fertilization (N-) on nitrogen fate in soil-plant system during 13-17 years in three experiments in Northern France. CC were present in all sites whereas tillage treatment and N fertilization rate were tested separately at one site. Crop biomass, N uptake and N leaching were monitored during the whole period. The N balance, i.e. the difference between N inputs and crop exportations, was only affected by fertilization rate whereas leached N varied with all techniques. CC was the most efficient technique to decrease N leaching (from 36 to 62%) and remained efficient on the long term. NT and N-had a positive but smaller impact. N storage in soil organic matter was markedly increased by CC (by 1024 kg ha(-1) yr(-1)), decreased by N- (-7.3 kg ha(-1) yr(-1)) and not significantly affected by NT. The differences in gaseous N losses (denitrification + volatilization) between treatments were assessed by nitrogen mass balance. CC establishment had no significant effect on N gaseous emissions while NT increased them by 3.6 +/- 0.9 kg N ha(-1) yr(-1) and N-reduced them by 13.6 +/- 4.6 kg N ha(-1) yr(-1). Catch crops appear as a win/win technique with respect to nitrate leaching and C and N sequestration in soil.