Soil aggregate loss during hillslope erosion processes directly reflects the interaction among the degree of soil aggregate breakdown, raindrop impact and runoff transportation. Different soil aggregate breakdown mechanisms result in the size distribution of soil aggregate is different. Furthermore, soil aggregate loss has an important influence on hillslope soil loss. However, existing studies have mostly focused on aggregate stability. Rainfall simulation experiments were conducted to investigate Mollisol aggregate loss during rainfall erosion processes. The experimental treatments included two rainfall intensities (50 mm/h and 100 mm/h), representative of erosive rainfalls in the black soil region of northeast China; two common slope gradients (5° and 7.5°); and two surface conditions (bare land and straw mulch cover). Each experimental treatment had two replications. The tested soil was the Mollisol, which was collected from the upper 20 cm of the plow layer in a maize field in Yushu City, Jilin Province, which is typical of the black soil of the region of northeast China. This study was conducted in the rainfall simulation laboratory of the State Key Laboratory of Soil Erosion and Dryland Farming on the Loess Plateau, Yangling City, China. A rainfall simulator system with a side-sprinkler was used to apply rainfall. The soil pan was 8 m long, 1.5 m wide and 0.6 m deep. During the rainfall simulations, runoff samples were collected in 15-L buckets. Half of each runoff sample was immediately processed through a set of sieves with apertures of 5 mm, 2 mm, 1 mm, 0.5 mm, and 0.25 mm. After sieving, the lost aggregate samples and the remaining half of each runoff samples were oven-dried to calculate the soil and aggregate losses. The results showed that sediment concentration in the bare land treatments was 27.5-141.3 times greater than that in the straw mulch cover treatments. The maximum values for sediment concentration were observed at the beginning of the rainfall simulation, in all treatments. This indicated that the dominant Mollisol aggregate breakdown mechanisms were slaking and mechanical breakdown, and that the effect of slaking was mainly exerted in the initial stages of rainfall. The < 0.25 mm micro-aggregates were the main aggregate size fraction lost in both the bare land and the straw mulch cover treatments. The < 0.25 mm micro-aggregate loss comprised 34.5-56.8% of the total aggregate loss in the straw mulch cover treatments, whereas the loss was > 82% in the bare land treatments. Compared with the bare land treatments, the loss of every size aggregate decreased > 33.3% in the straw mulch cover treatments; thus, the most significant differences in soil aggregate loss were observed in the ≥ 1 mm and < 0.25 mm size fractions between the bare land and the straw mulch cover treatments. Compared with the bare land treatments, the loss of the ≥1 mm size fraction and the < 0.25 mm micro-aggregates decreased 43.1%-96.4% and 99.0%, respectively, in the straw mulch cover treatments. The proportion of 0.25-2 mm aggregates in the sediment from the straw mulch cover treatments increased and was greater than that of the bare land treatments. These results indicated that raindrop impact was the main driving force for aggregate breakdown. The straw mulch cover eliminated the effect of raindrop impact on soil aggregate breakdown, and also decreased runoff transport capacity. The mean weight diameter (MWD) and geometric mean diameter (GMD) of aggregates in the sediments for the straw mulch cover treatments were 1.5-2.9 and 1.7-2.0 times greater than those for the bare land treatments. Compared with the bare land treatments, the mean weight soil specific (MWSSA) and fraction dimension (D) decreased 26.2%-32.9% and 5.1%-6.7%, respectively. The above four indicators reflect the aggregate loss characteristics of Mollisols.