Aquatic ecosystems have been identified as a globally significant source of nitrous oxide(N2O) due to continuous active nitrogen involvement, but the processes and influencing factors that control N2O production are still poorly understood, especially in reservoirs. For that, monthly N2O variations were monitored in Dongfeng reservoir(DFR)with a mesotrophic condition. The dissolved N2O concentration in DFR displayed a distinct spatial–temporal pattern but lower than that in the eutrophic reservoirs. During the whole sampling year, N2O saturation ranging from 144% to 640%, indicating that reservoir acted as source of atmospheric N2O. N2O production is induced by the introduction of nitrogen(NO3-, NH4+) in mesotrophic reservoirs, and is also affected by oxygen level and water temperature. Nitrification was the predominate process for N2O production in DFR due to well-oxygenated longitudinal water layers.Mean values of estimated N2O flux from the air–water interface averaged 0.19 μmol m-2h-1with a range of 0.01–0.61 μmol m-2h-1. DFR exhibited less N2O emission flux than that reported in a nearby eutrophic reservoir, but still acted as a moderate N2O source compared with other reservoirs and lakes worldwide. Annual emissions from the water–air interface of DFR were estimated to be 0.32×105mol N–N2O, while N2O degassing from releasing water behind the dam during power generation was nearly five times greater. Hence, N2O degassing behind the dam should be taken into account for estimation of N2O emissions from artificial reservoirs, an omission that historically has probably resulted in underestimates. IPCC methodology should consider more specifically N2O emission estimation in aquatic ecosystems, especially in reservoirs, the default EF5 model will lead to an overestimation.