[Objective] The distribution characteristics of rainfall erosivity R value in the yellow soil area of karst mountainous region in Central Guizhou Province were studied, in order to provide scientific references for regional soil erosion quantitative prediction, soil conservation planning and soil and water loss control. [Methods] Based on the data source of the daily rainfall record from 2013 to 2019, and the extracted data of rainfall process at 5 min intervals from 10 soil and water conservation monitoring stations in the karst yellow soil distribution area of Guizhou Province, the individual distribution characteristics, monthly distribution characteristics, inter-annual variation characteristics of R value, rainfall and rainfall intensity distribution characteristics were analyzed. [Results] ① The maximum R values was several times to a dozen times of the average R value of each monitoring stations. The percentage of the maximum R values in the corresponding year was also significantly different, and at least accounted for about 22.28% of the annual value. The contribution rate of several heavy rainstorms in a year was very high. ② The R value of rainfall erosivity was mainly distributed from April to September, and particularly concentrated from June to August. The R value from April to September accounted for about 90.00% of the annual R value in the monitoring stations, among which some stations were more than 95.00%. From June to August, the lowest R value accounted for 55.98% of the annual value, and the highest was 85.25%. ③ The average annual R value decreased significantly from southeast to southwest. The inter-annual variation coefficient (C_v) of R values showed an opposite trend, and increased significantly from southeast to northwest. The variation range of the inter-annual variation coefficient (C_v) of R values was 0.20～0.44, with a large interannual variation. ④ Moderate rain, heavy rain, rainstorm, and heavy rainstorm were the main rainfall erosivity levels, with which most stations accounting for more than 60.00%. Heavy rain, due to its relatively high frequency and long duration, contributed the most to the total R value. Overall, the frequency of heavy rainstorm was not high, but the value of rainfall erosivity R value of a single heavy rainstorm could be very large, which could change the R value distribution by a single heavy rainstorm. The peak area of R value distribution was 15—30 mm/h, and the average proportion is 31.97%. The occurrence of high rainfall intensity was more random, and the spatial distribution difference of the R value proportion generated by rainfall intensity greater than 60 mm/h was also large. [Conclusion] The spatial and temporal variation of rainfall erosivity R value is obvious in the yellow soil area of karst mountainous region in Central Guizhou Province, so it is necessary to take measures to prevent and control soil erosion according to local conditions.