A series of experiments was conducted to examine topographical variation, runoff erosion, and ion transport processes to explore the coupling relation of the different processes in slope evolution. The series includes 9 sequential experiments. In the first experiment, air-dried, crushed, and sieved loess was incrementally packed into a laboratory flume 5-m long and 1.5-m wide to a depth of 0.8 to 2.6 m. Copper nitrate (Cu(NO3)2)was mixed evenly in the top 0.2-m layer. The initial slope of the hillslope surface relief in the first experiment is 20 degrees and the initial slope of the latter experiments is the terminal relief of the former experiment. Simulated rain, applied at 90 mm/h for 90 min, was held constant in each experiment. The measured items in the experiment included slope elevation, sediment transport rate, and water discharge as well as dissolved and exchangeable nitrate (NO3-N) and copper ion (Cu2+). It was found that the sediment transport process was highly related to the topographical variation at different positions on the slope. Further analysis indicates that headcut erosion is highly related to the area draining to the rill head and landslip generally occurs at a bank with a gradient more than 65 degrees. However, the ion transport process is directly related to the processes of runoff and soil erosion rather than to the topographical variation process. Most NO3-N runs out with water, with infiltration at first and then with runoff, while the loss of Cu2+, which is affected by the pH of the solid-liquid system, is mainly determined by sediment transport owing to the different characteristics of NO3- and Cu2+. A series of experiments was conducted to examine topographical variation, runoff erosion, and ion transport processes to explore the coupling relation of the different processes in slope evolution. The series includes 9 sequential experiments. In the first experiment, air-dried, crushed, and sieved loess was incrementally packed into a laboratory flume 5-m long and 1.5-m wide to a depth of 0.8 to 2.6 m. Copper nitrate (Cu (NO3) 2) was mixed evenly in the top 0.2-m layer. The initial slope of the hillslope surface relief in the first experiment is 20 degrees and the initial slope of the latter experiments is the terminal relief of the former experiment. Simulated rain, applied at 90 mm / h for 90 minutes, was held constant in each experiment. measured items in the experiment included slope elevation, sediment transport rate, and water discharge as well as dissolved and exchangeable nitrate (NO3-N) and copper ion (Cu2 +). It was found that the sediment transport process was highly related to the topographical variation at different positions on the slope. Further analysis indicates that headcut erosion is highly related to the area draining to the rill head and landslip generally occurs at a bank with a gradient more than 65 degrees. However, the ion transport process is directly related to the processes of runoff and soil erosion rather than to the topographical variation process. Most NO3-N runs out with water, with infiltration at first and then with runoff, while the loss of Cu2 +, which is affected by the pH of the solid -liquid system, is mainly determined by sediment transport due to the different characteristics of NO3- and Cu2 +.