A field experiment was carried out to investigate the effects of different rice-rape rotation systems on methane and nitrous oxide emissions, which were measured using the static chamber/gas chromatography method, prediction of their global warming potentials (GWP), and greenhouse gas intensity (GHGI) in paddy fields. The results showed that the average cumulative emissions of CH4from a double cropping paddy field, single season rice field (including middle or late), rape field, and leisure land were 135.25, 55.64, 5.05, and 1.89 kg ·hm-2, respectively. The CH4 emission during the rice season accounted for 91.8%-98.5% of the annual CH4 emission in different rotation years, and the contents of dissolved organic carbon in paddy soil exhibited a significantly positive correlation with CH4 emission. The CH4 emission in conventional late rice paddy was 18.7% higher than that of hybrid late rice paddy (P<0.05). The average cumulative emissions of N2O from double cropping paddy field, single season rice field (including middle or late), rape field, and leisure land were 0.94, 0.64, 1.38, and 0.24 kg ·hm-2, respectively. Out of the total annual N2O emission, 57.2% to 70.2% was from the rape field; 17.8% and 30.6% was due to the winter fallow treatments with previous crop type of double corpping rice and single season rice, respectively. There was no significant difference in N2O emission between hybrid rice and conventional rice paddy fields. The GWP of double cropping rice-winter fallow and double cropping rice-rape was higher than that of rice-rape and rice-winter fallow, and the GWP of CH4 in rice season accounted for 71.2% to 90.9% of the annual GWP of rotation. The highest treatment of GHGI was rice-rice-rape, and the treatment of rice-oil and rice-winter slack was lower. According to the comprehensive environmental and economic benefits, the late hybrid rice-rape patterns should be selected to reduce greenhouse gases in multiple cropping rice fields of South China.