The precipitation is the key process in water cycle. It directly affects temporal and spatial distribution of runoff process and amount of available water resources. To investigate space -time evolution of precipitation in typical small watersheds of southern China under climate change, this paper analyzed the annual precipitation, flood season and non-flood season precipitation, annual maximum daily precipitation and annual maximum monthly precipitation by using the nonparametric statistics including Mann Kendall trend test method and Mann-WhitneyPettitt verification method based on the 47-year observed daily precipitation from 10 stations in the Binjiang River Basin, a tributary of the Beijiang River in the Pearl River Basin. The result shows that the multi-year average precipitation of the northern part in the basin is less than that in the south, which has close relation to regional airflow direction and valley terrain. At the same time, general decline of annual precipitation, including southern basin has a significant decline and the northern part shows no significant decline, which is mainly the precipitation in the flood season (April-September ) reduced, while the precipitation in non-flood season (October-March in the next year) changed little. Both annual maximum daily precipitation and annual maximum monthly precipitation have a trend of decrease. The annual precipitation and flood season precipitation in the southern part of the basin had a significant variation in 1983, while in the north, the annual maximum daily precipitation and annual maximum monthly precipitation have respectively significant variation in 1987 and 1985.