Abstract: In the North China Plain, a field experiment was conducted to measure nitrous oxide (N2O) and methane (CH4) fluxes from a typical winter wheat–summer maize rotation system under five integrated agricultural management practices: conventional regime [excessive nitrogen (N) fertilization, flood irrigation, and rotary tillage before wheat sowing; CON], recommended regime 1 (balanced N fertilization, decreased irrigation, and deep plowing before wheat sowing; REC-1), recommended regime 2 (balanced N fertilization, decreased irrigation, and no tillage; REC-2), recommended regime 3 (controlled release N fertilizer, decreased irrigation, and no tillage; REC-3), and no N fertilizer (CK). Field measurements indicated that pulse emissions after N fertilization and irrigation contributed 19–49 % of annual N2O emissions. In contrast to CON (2.21 kg N2O-N ha−1 year−1), the other treatments resulted in significant declines in cumulative N2O emissions, which ranged from 0.96 to 1.76 kg N2O-N ha−1 year−1, indicating that the recommended practices (e.g., balanced N fertilization, controlled release N fertilizer, and decreased irrigation) offered substantial benefits for both sustaining grain yield and reducing N2O emissions. Emission factors of N fertilizer were 0.21, 0.22, 0.23, and 0.37 % under CON, REC-1, REC-3, and REC-2, respectively. Emissions of N2O during the freeze–thaw cycle period and the winter freezing period accounted for 9.7 and 5.1 % of the annual N2O budget, respectively. Thus, we recommend that the monitoring frequency should be increased during the freeze–thaw cycle period to obtain a proper estimate of total emissions. Annual CH4 fluxes from the soil were low (−1.54 to −1.12 kg CH4-C ha−1 year−1), and N fertilizer application had no obvious effects on CH4 uptake. Values of global warming potential were predominantly determined by N2O emissions, which were 411 kg CO2-eq ha−1 year−1 in the CK and 694–982 kg CO2-eq ha−1 year−1 in the N fertilization regimes. When comprehensively considering grain yield, global warming potential intensity values in REC-1, REC-2, and REC-3 were significantly lower than in CON. Meanwhile, grain yield increased slightly under REC-1 and REC-3 compared to CON. Generally, REC-1 and REC-3 are recommended as promising management regimes to attain the dual objectives of sustaining grain yield and reducing greenhouse gas emissions in the North China Plain.