Abstract: Splash erosion is the initial stage of soil erosion by water, particularly an important process of hillslope erosion. The main cause of splash erosion can be resulted from the destructive force of rain drops, when acting on soil surface aggregates. As such, the splash erosion strongly depends on the rainfall intensity and slope gradient. Most previous studies focused on the characteristics of splash erosion, and dominated affecting factors, in order to reveal the specific mechanism of splash erosion. Moreover, most current studies on splash erosion were usually conducted on non-ridge slope in the farmland. However, little information is available on the up-down ridged slope, especially for the distribution of soil water stable aggregates during splash erosion process in the typical black soil region of Northeast China. Therefore, in this study, an in-situ field simulated rainfall experiment was conducted to investigate the effects of rainfall intensity and slope gradient on the characteristics of splash erosion, and the distribution of soil water stable aggregates in the process of splash erosion. Three rainfall intensities (30, 60, 90 mm/h) of representative erosive rainfall, and two slope gradients (5° and 10°) were set, where all treatments were replicated three times. The soil in this experiment was a Mollisol (USDA System of Soil Taxonomy), with 31.6% sand (> 50 μm), 30.8% silt (50 to 2 μm), and 37.6% clay (< 2 μm). The filed study was conducted in the Hailun Monitoring and Research Station of Soil and Water Conservation, Chinese Academy of Sciences (47°21′16.95″N，126°49′56.43″E), located at the Hailun city, Heilongjiang Province, the center of the typical Mollisols zone in Northeast China. An up-down ridge system was used, where the ridged interval on the farmland was 70 cm, and the vertical height difference between ridge and furrow was 20 cm. In addition, the splash board was applied to measure splash erosion. The results showed that the total splash erosion increased by 2.5 to 17.9 times, as the rainfall intensity increased from 30 mm/h to 90 mm/h, whereas, the total splash erosion significantly increased by 30.65% to 64.34% (P < 0.05), as the slope gradient increased from 3° to 5°. The relationship between splash erosion with rainfall intensity and slope gradient can be expressed as a power function. Besides, at the rainfall intensities of 30 mm/h and 60mm/h, the splash erosion rate first rapidly decreased, then gradually decreased, and finally reached a steady state. At the rainfall intensity of 90 mm/h, the splash erosion rate first rapidly increased to the maximum, then rapidly decreased, and finally reached a steady state. Furthermore, the <1mm soil aggregate was the main fraction of aggregate size that detached by raindrop splash, accounting for 79.01% of total splash erosion. Specifically, the proportion of 0.5-1 mm soil aggregate was the most, accounting for 32.94% of total splash erosion, while, the proportion of 2-5 mm soil aggregate was the least, accounting for 3.36% of total splash erosion. In addition, at the rainfall intensities of 30 mm/h and 60mm/h, the <0.25 mm and <2 mm soil aggregates reached the steady state at the end of rainfall, respectively. The proportion of micro aggregates first rapidly increased, then slowly increased, and finally reached a relatively stable stage. At the rainfall intensity of 90 mm/h, the 1-5 mm and <0.25 mm soil aggregates showed a linear trend, whereas, the proportion of micro aggregates showed a decrease trend. This study can provide a sound scientific basis to control water erosion in the typical black soil region of Northeast China.