Abstract: Rill erosion caused by concentrate flow is one of the main erosion types on cultivated slope in the Loess Plateau. It is necessary to research on the response of concentrate flow hydrodynamic characteristics for a better understanding of rill erosion mechanism. However, the optimal runoff hydrodynamic parameter for estimating detachment rate was still ambiguous. An indoor concentrate scouring experiment was carried out was carried out in the State Key Laboratory of Soil Erosion and Dryland arming on the Loess Plateau, Institute of Soil and Water Conservation, Chinese Academy of Sciences and Ministry of Water Resources, China to investigate the response of runoff hydrodynamic characteristics to detachment rate under concentrate flow condition with different inflow rates and slope gradients. Loessal soil collected from Yanan (35°21′-37°31′ N and 107°41′-110°31′ E) in Shaanxi province, a kind of typical soil in the Loess Plateau, was prepared for this research. The experiments were applied to a soil plot with 5 m long, 1 m wide and 0.5 m deep. Packing was carried out layer by layer to attain the desired uniform bulk density (about 1.25 g/cm-3) with 40 cm in depth. The bottoms of the boxes were perforated and covered with a layer of 10 cm sand under the gauze to facilitate even drainage of percolating soil water. After packing, the soil was watered to saturation with an electric sprayer to reduce the variability caused by packing. Four flow rate (10, 15, 20 and 25 L/min) combined with four slope gradient (10°, 15°, 20° and 25°) were designed for this research. The experiment lasted for 10 min after runoff initiation. Runoff and sediments were collected in a series of plastic containers at intervals of 1 min throughout the 10 min. The volume of water in each container was measured, and the sediment was dried in an oven and weighed. The flow velocity was measured by dye-tracing technique within the 1-4 m away from the bottom and the flow width was also measured at 4 sections between 0.5-4.5 m to estimate flow depth at every minute during the experiment. The relations between runoff hydrodynamic characteristics, including shear stress, stream power, unit stream power and unit energy of water-carrying section, and detachment rate were analyzed. The results showed that all the mean and instantaneous runoff hydrodynamic characteristics factors fitted the detachment rate well with different regressions equations except instantaneous unit energy of water-carrying section. The mean runoff hydrodynamic characteristics factors were better than those of average values for fitting with detachment rate. The optimal runoff hydrodynamic characteristics factor in our research was mean stream power because it was of the largest determination coefficient 0.97. The curve of linear regressions of mean shear stress and stream power with detachment rate became ascended because of collapse during the experiment process, which also led to a negative value for corresponding critical shear stress and stream power. By comparing results with that from a published paper that only considered the flow effect on soil surface in the same soil, the detachment rate directly estimated based on mean stream power were more reasonable than those estimated based on mean shear stress. The collapse could account for 90.93% of the detachment rate, indicating an important role of collapse during rill development process. The results provide valuable information for a better understand of the response of concentrate flow hydrodynamic characteristic factors to detachment rate and its corresponding erosion mechanism.