The return of rice straw and leguminous green manures to soil are two effective ways of improving soil fertility in south China. Despite the popularity of this practice, our understanding of the underlying processes is relatively poor. In this study, rice straw (C/N 63:1) and green manure (hairy vetch, C/N 14:1) were incubated alone or in mixtures (at different C/N ratios of 25 and 35) with soil to evaluate the effects on soil N availability and C or N loss in a paddy soil under waterlogging. All residue treatments significantly enhanced CO2 and CH4 emissions, but decreased N2O emissions. Soil dissolved organic C and N, microbial biomass C, upper aqueous layer soluble organic C and N, and mineral N were also enhanced in residue-treated soils (except the rice straw treatment), while soil microbial biomass N and mineral N were lower in the rice straw treatment. Changes in plant residue C/N ratios, dissolved organic C/N ratios and cellulose significantly affected soil greenhouse gas emissions and soil activated C and N concentrations. Additionally, the treatment with green manure alone had the largest C and N losses; incorporation of residue mixture with a C/N ratio of 35 caused the largest net global warming potential (nGWP) among the amended soils. These results indicated that co-incorporation of rice straw and green manure can alleviate the limitation brought by only applying rice straw (leading to N immobilization) or the sole application of leguminous green manure (causing high C and N losses); the residue mixture with a C/N ratio of 25 was better because of its lower nGWP.