Abstract: Plant litter has been accumulated greatly along with vegetation restoration on the Loess Plateau. Besides covering soil surface, plant litter can be incorporated into topsoil under natural circumstances by soil splash, sediment deposition, and soil-dwelling animal activities. The distribution (covering soil surface or incorporated into surface soil) and the decomposition of plant litter can modify soil physical properties (including soil detachment capacity and soil erosion resistance) which are closely related to soil detachment process by overland flow. This study was conducted to investigate the effect of vegetation restoration on soil detachment process for black locust (Robinia pseudoacacia L.) stand with different ages of returning farmland on the Loess Plateau. The experiments were performed in the Zhifanggou small watershed (36°46'28"-36°46'42"N, 109°13'46"-109°16'03"E) in August, 2015. After a complete watershed survey, altogether 6 sampling sites were selected, including 1 corn (Zea mays L.) field and 5 black locust stands of 10, 15, 20, 30, and 40 years. Soil samples were collected from the top 5 cm soil layer using a circular steel ring with 10.0 cm diameter and 5.0 cm depth. Soil detachment was measured in a 4.0 m long, 0.35 m wide flume. Prior to the test of soil detachment capacity, soil samples were saturated in a container for 8 h and drained for 12 h. Then the soil samples were inserted into a hole (0.5 m away from the outlet of flume) on the flume bed and were scoured by flowing water under designed flow shear stress. A total of 180 soil samples were collected from different aged black locust stands and scoured under 6 flow shear stresses. In this study, 6 combinations of flow rates and slopes were applied to obtain 6 different flow shear stresses (5.4-17.4Pa) with similar intervals. Rill erodibility is defined as the increase in soil detachment capacity per unit increase in flow shear stress. Critical shear stress is a threshold parameter defined as the value above which a rapid increase in soil detachment capacity per unit increase in shear stress occurs. When the measured detachment capacity was plotted against the flow shear stress, rill erodibility and critical shear stress could be determined from the slope of the fitted straight line and its intercept on the X axis. The results showed that the density of plant litter in 0-5 cm surface soil was 0.07, 0.18, 0.32, 0.29 and 0.17 kg/m2 for 10, 15, 20, 30, and 40-year-old black locust stands, respectively, with the mean value of 0.20 kg/m2. The 40-year-old black locust stands had more capillary porosities, soil organic matter, and water stable aggregates, and smaller soil bulk density. Soil detachment capacity decreased by 49.8%, 73.9%, 85.0%, 86.0%, and 87.0% for 10, 15, 20, 30, and 40-year-old black locust stands, respectively, compared to the control. The soil detachment capacity decreased significantly over time as an exponential function (R2=0.97, P=0.006). The control (corn field) had the biggest rill erodibility (0.29 s/m), and the 40-year-old black locust stand had the smallest rill erodibility (0.04 s/m). Compared to the control, the rill erodibility for 40-year-old black locust stand was reduced by 86.3%. The black locust stand with 20 years had the biggest critical shear stress (4.78 Pa), and the control had the smallest critical shear stress (4.15 Pa). The critical shear stress increased by 10.1% for 40-year-old black locust stand compared to the control. The critical shear stress increased significantly with litter density as a linear function (R2=0.67, P=0.046). Rill erodibility was better than critical shear stress in the respect of reflecting the variation of soil detachment. The soil-plant litter system has a larger erosion resistance to flowing water than the control. Importantly, the soil-root system may have a larger structural stability and erosion resistance to flowing water than soil-plant litter system.