Few investigations have been made on the impact of elevated ozone (O3) concentration on methane (CH4) emission from rice paddies. Using open-top chambers in situ with different O3 treatments, CH4 emissions were measured in a rice paddy in Yangtze River Delta, China in 2007 and 2008. There were four treatments applied: charcoal-filtered air (CF), nonfiltered air (NF), and charcoal-filtered air with different O3 additions (O3-1 and O3-2). The mean O3 concentrations during the O3 fumigation were 19.7, 22.6, 69.6 and 118.6 ppb in 2007 and 7.0, 17.4, 82.2 and 138.3 ppb in 2008 for treatments CF, NF, O3-1 and O3-2, respectively. The rice yields, as compared with CF, were reduced by 32.8% and 37.1%, 58.3% and 52.1% in treatments O3-1 and O3-2 in 2007 and 2008, respectively. The diurnal patterns of CH4 emission varied temporally with treatments and there was inconsistence in diurnal variations in CH4 emissions from the paddy field. The daily mean CH4 emissions were significantly lower in treatments O3-1 and O3-2 than those in treatments CF and NF. Compared with CF treatment, CH4 emissions from the paddy field were decreased to 46.5% and 38.3%, 50.6% and 46.8% under treatments O3-1 and O3-2 in the whole growing seasons of 2007 and 2008, respectively. The seasonal mean CH4 emissions were negatively related with AOT40 (accumulative O3 concentration above 40 ppb; P < 0.01 in both years), but positively related to the relative rice yield (reference to CF; P < 0.01 in 2007 and P < 0.001 in 2008), aboveground biomass (P < 0.01 in both years) and underground biomass (P < 0.01 in 2007 and P < 0.05 in 2008). The decreased CH4 emission from the rice paddy due to an increased O3 exposure might partially mitigate the global warming potential induced by soil carbon loss under elevated O3 concentrations.