Using the daily precipitation data of the long-term series of meteorological stations on the Qinghai-Tibet Plateau, the percentile threshold method is used to determine the extreme precipitation threshold, calculate the extreme precipitation index and analyze its spatial and temporal distribution characteristics, in order to provide reference for regional climate change prediction and disaster prevention and mitigation countermeasures. The results show that: (1) From 1961 to 2017, the annual precipitation of Qinghai-Tibet Plateau showed an upward trend, with a rate of 8.06 mm/10 a, and the average annual precipitation reached 472.36 mm. The minimum precipitation tendency rate of 78 stations is -25.46 mm/10 a, and the maximum value is 43.02 mm/10 a. The precipitation of 15.38% of the stations is decreasing, which is mainly distributed in the east and south of the plateau, and the precipitation of the remaining 84.62% of the stations is increasing. (2) The average threshold value of extreme precipitation in the Qinghai-Tibet Plateau is 23.11 mm, with error values ranging from 7.84 mm to 51.90 mm. The high value centers are located in Gongshan and Muli of Hengduan Mountains, while the low value centers are located in the northern flank of Qaidam Basin and Kunlun Mountains. (3) The extreme precipitation, the number of days of extreme precipitation and the contribution rate of extreme precipitation at all the stations in the Qinghai-Tibet Plateau show an obvious upward trend. Although the intensity of extreme precipitation is also rising, the trend is not obvious, which shows that the increase of extreme precipitation in the plateau is not caused by the intensity of extreme precipitation, but by the increase of the frequency of extreme precipitation. Although the extreme precipitation and days of extreme precipitation in the Qaidam Basin do not show a high value level, the contribution rate of extreme precipitation is larger, which suggests that although there is less precipitation, extreme precipitation events frequently occur in this area.