ABSTRACTThe duration of rainfall is a significant factor in the process of splash erosion. Nevertheless, there is a lack of comprehensive documentation regarding the extent and characteristic of soil aggregate disintegration resulting from raindrop impact during splash erosion. To examine the properties of the aggregate disintegration mechanism during various of splash erosion, three soil samples (S‐1, S‐2, and S‐3 labeled as the organic matter was 19.77, 31.70, and 26.83 g kg−1) were employed to simulated splash erosion under six durations of rainfall (5, 10, 15, 20, 30, and 40 min). Aggregates ranging in size from 1 to 5 mm were utilized in the preparation of test soils for different antecedent moisture contents (3%, 5%, 10%, and 15%). The result demonstrated that the slaking effect, as observed in the fast wetting of the Le Bissonnais method, is attributed to the sudden release of air entering the water, irrespective of the water immersion duration. Upon increasing moisture content, under fast wetting conditions, the mean weight diameter (MWD) values for S‐2 and S‐3 samples increased from 1.25 and 1.31 to 1.85 and 1.61, respectively. Subsequently, at a moisture content of 10%, the MWD decreased to 1.84 and 1.49 for S‐2 and S‐3 samples, respectively. In contrast, S‐1 sample exhibited an increasing trend from 0.34 to 0.93. The sensitivity of aggregates to disintegration from slaking (RSI) decreased by 0.58, 0.43, and 0.29, while mechanical impact (RMI) decreased by 0.23, 0.11, and 0.08 with increasing moisture content, respectively. The rate of splash erosion for the S‐1 sample exhibited an initial increase followed by a subsequent decrease as the duration of rainfall increased. The S‐1 sample, with the lowest organic matter content, exhibited the highest content of microaggregate content (< 0.25 mm) of splashed particles (referring to soil particles transported out of the splash area by raindrops), ranging from 31.62% to 65.57%. The content of macroaggregate (> 0.25 mm) in the residual soil (indicating soil particles that were left in the area affected by splashing) exhibited a gradual reduction as the duration of rainfall increased. As the soil reached saturation, the influence of physicochemical dispersion became more pronounced with the prolonged duration of rainfall. The forces of slaking and physicochemical dispersion predominantly disrupted > 1 mm aggregates, while the destruction of aggregates within the 0.25–1 mm range was insufficient. This study is valuable for understanding the splash erosion process and provides a scientific foundation for the improvement of soil erosion prediction models.