Abstract: Based on the observed daily precipitation during 1965-2014 at 130 meteorological stations of the Yangtze river basin, the regional frequencies and spatio-temporal characteristics of extreme precipitation were analyzed by using the L-moments approach, statistical tests and spatial analysis techniques. The results are as follows. (1) The annual maximum 1-, 3-, 7- and 10-day precipitation sequences of the Yangtze River basin can be divided into seven homogeneous sub-regions by fuzzy c-mean classification together with heterogeneity test. The results of Goodness-of-fit measure showed that the GEV and GNO distributions are the optimal distributions for most of the regional extreme precipitation sequences. (2)The accuracy of the extreme precipitation growth curve and quantile estimates was evaluated by using Monte Carlo simulations with inter-station dependence, and the results showed that the root-mean-square errors (RMSE) were bigger and the 90% error bounds were wider with inter-site dependence than those without inter-site dependence for both the regional growth curve and quantile estimates. (3) The regional growth curves with 90% error bands for each homogeneous region showed that the quantile estimates are reliable enough for return periods of less than 100 years, which indicated that extreme precipitation events are highly probable to occur in the middle and lower Yangtze river basin and Szechwan Basin, and hence higher risk of floods. (4) The spatial patterns of annual extreme daily precipitation with return period of 100 years indicated that extreme rainfall increased gradually from the upper reaches to the lower reaches of the Yangtze River, showing higher risks of floods in the middle and lower reaches of the Yangtze River, which was consistent with the regional growth curves.