To identify the key soil parameters influencing CH₄ emission from rice paddies, an outdoor pot experiment with a total of 18 paddy soils was conducted in Nanjing Agricultural University during the 2000 rice growing season. The seasonal average rate of CH₄ emission for all 18 soils was 6.42±2.70 mg m^–2 h^–1, with a range of 1.96 to 11.06 mg m^–2 h^–1. Correlation analysis indicated that the seasonal average of CH₄ emission was positively dependent on soil sand content and negatively on total N as well as NH₄ ⁺-N determined before rice transplanting. Copper content of soils had a significant negative impact on CH₄ emission. No clear relationship existed between CH₄ emission and soil carbon content. In addition, soil type cannot explain the variability in CH₄ emission. Soil parameters influencing CH₄ emission were different as rice growth and development proceeded. A further investigation suggested that the seasonal average rate of CH₄ emission could be quantitatively determined by a linear combination of soil NH₄ ⁺-N, available copper, the ratio of available to total sulphur, and the ratio of available to total iron. Moreover, the average rates of CH₄ emission in the vegetative, reproductive and ripening stages could be also respectively described by a linear combination of different soil variables.