Abstract: Abstract: Soil detachment is the initial stage of soil erosion, which refers to the process of soil surface particles detached from the original soil under the action of rainfall splash or runoff, and is the main source of erosion sediment. The slope after freeze-thaw provides a large number of effective materials for water erosion. In recent decades, as the global climate tends to be warmer, the effects of freeze-thaw in the areas of high latitude and high elevation have been intensified. This study was conducted to investigate the effects of freeze-thaw on soil detachment capacity and erosion resistance by means of indoor simulation freeze-thaw and runoff scour tests. Soil samples were collected from the top 20 cm of abandoned land in Loess Plateau of China. To remove stones, grass, and other debris, air-dried soil samples were sieved through a 5 mm mesh. Soil samples were stored in polyvinyl chloride (PVC) cylindrical boxes based on the bulk densities in the field. Then the samples were frozen at -10 ℃ for 12 hours and thawed at room temperature between 5℃ and 10℃ for 12 hours to simulate the natural phenomenon of night freezing and day thawing. Soil detachment capacity was measured in a 4.0 m long, 0.15 m wide flume. The loessal soil was subjected to 0, 1, 5, 10 freeze-thaw cycle times before it was scoured, while the slope and flow discharge of flume experiments were 10°, 15° and 12, 18, 24 L/min, respectively. The results showed that slope, flow discharge and freeze-thaw cycle times all had significant effects on soil detachment capacity (P<0.05), with contribution rates of 17.94%, 19.96% and 18.43%, respectively. Soil detachment capacity basically increased with the increase of slope and flow rate. The mean value after freeze-thaw (5.28±2.48 g/(cm2·min)) was significantly higher than that before freeze-thaw (2.39±1.71 g/(cm2·min)), but the degree of increase after freeze-thaw was significantly lower than that before freeze-thaw. Under different slope and flow rate conditions, soil detachment capacity increased significantly after first freeze-thaw (P<0.05). The variation trend of soil detachment capacity with the times of freeze-thaw cycles was significantly different. Only when the slope and flow rate were both small (10° and ≤18 L/min) or large (15° and ≥18 L/min), the trend increased significantly. After 1, 5 and 10 times of freeze-thaw cycle, the rill erodibility increased by 1.25, 1.66 and 1.72 times respectively, and gradually became stable with the increase of freeze-thaw cycle times. The critical shear stress decreased significantly after freeze-thaw cycles and had no significant relationship with freeze-thaw cycle times. After freeze-thaw, the means of soil bulk density, water stable aggregates and shear strength decreased by 6.61%, 24.77% and 21.35%, respectively. Under the condition of freeze-thaw, rill erodibility was negatively correlated with water stable aggregates and shear strength, but positively correlated with soil porosity. This study can provide reference for the study of complex erosion of freeze-thaw and water.