As an important supplement to the ground observation data, satellite precipitation products are of particularly important for areas lack of data or without data. The Niyang River Basin in the arctic-alpine mountainous region of China is a typical area that of data shortage. Based on the precipitation data of the five meteorological stations from 2001 to 2016 and the China ground precipitation daily grid point dataset from 2001 to 2013, the inverse distance weighted interpolation method was employed to process the precipitation data. The accuracy of TRMM precipitation data was estimated by correlation coefficient (R) and relative error (BIAS) at monthly and daily scales in the Niyang River Basin. The results showed that the monthly precipitation data of TRMM and the measured data had good consistency in the whole basin (R=0.90), but the consistency was poor and BIAS was large for the daily precipitation data (R=0.29). For the single station, the accuracy of TRMM monthly precipitation data was high (R>0.9), while the R of the daily scale were less than 0.5 and the precision was low. On the spatial scale, the spatial distribution of the accuracy of TRMM precipitation was obtained by Kriging interpolation method. The accuracy of monthly precipitation decreased from west to east, while the accuracy of daily precipitation illustrated an opposite trend. Based on the high precision of TRMM monthly precipitation data, the temporal and spatial distribution characteristics of precipitation in the Niyang River were analyzed. The results showed that most of the precipitation in the Niyang River basin was concentrated from May to September, and less precipitation happened from November to February. The annual rainfall showed an increasing trend from northwest to east, and the precipitation varied greatly in different regions.