Abstract: Rainfall is the main driving force of non-point source pollution diffusion, and it is also one of the key factors to be considered in the control of non-point source pollution. However, natural rainfall is uncertain. In order to determine the experimental conditions required for the research on non-point source pollution caused by rainfall, the rainfall intensity and its variation range in three typical demonstration areas in north and south China were taken as parameters. Using an artificial rainfall simulator with simulated rainfall intensity of 15-120 mm/h, the simulated rainfall processes under 10 rainfall intensities (15, 20, 25, 30, 40, 45, 50, 60, 90 and 120 mm/h) and 2 rainfall heights (2, 4 m) were studied. The accuracy of rainfall intensity, uniformity of rainfall, and the size and distribution of raindrop diameter were analyzed. Results showed that the rainfall uniformity of the device was basically maintained above 80% in 4 m of rainfall height with the rainfall intensity of 15-120 mm/h, and rainfall height of 2 m could be maintained at 60%. The rainfall intensity distribution was large in the center and small in the periphery, with a divergence from the center to the surrounding areas. The median diameter of a raindrop was 0.85-2.28 mm, and the maximum raindrop diameter was not more than 3.02 mm. The artificially simulated rainfall parameters were similar to those of natural rainfall. The rainfall parameters at a height of 4 m were better than the same parameters at a height of 2 m when the rainfall intensity ranged from 15-35 mm/h, which was suitable for the study of non-point source pollution in the northern crop growing area, such as corn and sunflower. The study on rainfall at a height of 2 m was more suitable for the southern rice growing area. The results of this study lay a foundation for exploring the mechanism of rainfall release of agricultural non-point source pollution in China.