[Objective] This study aimed to investigate the effect of direct-seeding with non-flooding and wheat residue returning patterns on greenhouse gas emission from rice paddy. [Method] Two rice cultivars currently used in the production, Yangdao 6 （an indica） and Yangjing 4038 （a japonica）, were field grown using a direct-seeding method, and four treatments, wheat straw incorporation into soil and traditional flooding （SlF）, non-flooding and wheat straw mulching （NSM）, non-flooding and wheat straw incorporation into soil （NSl） and traditional flooding （no straw returned, Control, TF）, were imposed after sowing to maturity. Effects of direct-seeding with non-flooding and wheat residue returning patterns on CH4, N20 and CO2 emissions were investigated by using the method of static chamber-gas chromatographic tech- niques. [Result] Grain yield showed no significant difference between non-flooding and flooding treatments, but was significantly higher under the SlF than under any other treatments. The emission flux of CH4 and CO2 under TF and SlF exhibited a single peak curve, while changed little under the NSl and NSM The emission flux of N2Oshowed multiple perk curves for all the treatments. Compared with TF, SlF significantly increased mean emission flux of CH4 or N2O, decreased emission of N20, while NSl and NSM significantly decreased the mean emission flux of OH4, and increased emission flux of N2O and CO2. SIF also increased Green Warm Potential （GWP） of CH4, N2O and CO2 and the GWP per unit grain yield by 47.3%- 53.7% and 32.2%-39.4%, respectively. Both NSl and NSM decreased GWP by 24.2%-29.6% and 30.1%-35.5%, and the GWP per unit grain yield was decreased by 21.7-27.2% and 25.6%-31.1%, respectively. [Conelusion] both NSl and NSM could significantly reduce greenhouse effect of CH4, N2O and CO2 meanwhile maintain a high grain yield.
