An empirical orthogonal function (EOF) analysis of monthly TRMM/PR (precipitation radar) data set during 1998-2008 was used to examine dominant patterns of precipitation variability in the Yangtze River Basin attempting to explain variation revealed by standard deviation (Std) and coefficient of variation (CoV). The Std and CoV results have represented precipitation variation in different emphasis. The EOF analysis based on the TRMM/PR anomaly data shows that patterns of large scale precipitation variation can be well identified. A primary analysis reveals that the physical meanings of the first several leading EOFs are the general circulation of atmosphere and the prevailing winds, which account for the majority variance in precipitation. The leading EOFs explain the Std characters satisfyingly, but have not provided an interpretation for CoV analysis due to input data selection.