[Objective] This study aimed to testify the method of estimating the real precipitation with TRMM data with the aid of the observed precipitation and TRMM precipitation from 2002 to 2011.[Methods] The applicability of TRMM precipitation data in Changsha-Zhuzhou-Xiangtan region of Hunan Province from 2002 to 2011 was analyzed based on the precipitation data from 12 rainfall observation stations with the methods of correlation coefficient, error analysis, relative deviation analysis and statistical analysis.[Results] The fitting degree of TRMM data and observation station data in these regions were good. The goodness-of-fit of yearly rainfall was 0.69 and the ones of the spring, summer, fall and winter respectively were 0.68, 0.52, 0.66 and 0.34. The goodness-of-fit of monthly rainfall reached to 0.73 and that of the every station was good too. Methods of relative deviation analysis and error analysis indicated that the TRMM rainfall data was generally lower than the observed precipitation, which was mainly for that the area representativeness of the signal observational station was far smaller than that of the TRMM. This was because of that the cities' rainfall extensity was stronger than rural area caused by urbanization and that the rainfall observation stations are all located in the edge of cities. The deviation has some definite relations with typographic relief. The more the topographic relief is, the greater the deviation will be.[Conclusion] The TRMM rainfall data has some definite applicability in the region of Changsha-Zhuzhou-Xiangtan region. We can adopt it for different uses but we should have some indispensible correction with observational data. The TRMM data can also be a reference for the use of the limited previously rainfall observation data in order to improve the spatial accuracy.