Based on the gauge stations hourly precipitation datasets ,comprehensively assessed the performance of two multi-satellite precipitation products (GSMaP_MVK, GSMaP_Gauge) on an extremely heavy rainfall event in May 2010 in Guangxi.The results showed that: 1,The spatial distribution patterns of the precipitation can be well captured by two satellite precipitation products,the correlation coefficient between them and the ground observation data are 0.73 and 0.75,respectively.2,Over the Guangxi region,compared with the ground observation data, GSMaP_MVK and GSMaP_Gauge overestimated the precipitation by 24.34% and 45.88% respectively.At the center of the storm, two satellite precipitation successfully captured the temporal characteristics of precipitation and underestimated the total precipitation at the same time, the underestimate rates were 17.50% and 26.93%, respectively.3,After ground-based correction,GSMaP_Gauge reduced the miss probability of rain events and produced more excess light rain events。4,Although the two satellite products underestimated heavy rainfall to some degree,both of them can predict the occurrence of debris flow.