Based on the simulation of 1 000-year control experiment by the Community Earth System Model (CESM) with constant pre-industrial external forcing, 212 El Niño events and 226 La Niña events were identified according to the monthly SST (Sea Surface Temperature) of Nino3.4 region. Furthermore, the spatial pattern of May to September precipitation in eastern China in the El Niño (La Niña) episodes and their following years were illustrated. The relationship between the ENSO episodes and the floods/droughts of four regions including North China, Jianghuai, Jiangnan and South China in eastern China were analyzed. The results show that during the El Niño episodes, precipitation decreases by 2%-10% (with respect to the mean value of the whole 1 000 years) in North China and South China from May to September, and precipitation slightly increases by 0-2% in the Yangtze River Basin. In the following year of El Niño, precipitation in Jiangnan region turns to increase by 2%-10%, but precipitation in North China is still in decreasing conditions. During the La Niña episodes, precipitation in North China increases by 2%-10%. In the following year of La Niña, precipitation in Jianghuai decreases obviously by 2%-5%. The stronger El Niño (La Niña) events lead to more intense precipitation changing, and the percentages of precipitation anomalies increase. In the year when El Niño hasn’t vanished and La Niña starts to develop, the probability of flooding in the Jiangnan region might be more than twice as that in normal years. This finding provides background of abnormal climate for understanding the 2016 extreme flood in Yangtze River, and thoroughly reveals the influence of inter-annual internal variability in climate system on the changes of precipitation’s spatial patterns and regional flood/drought.