Soil erosion, which includes soil detachment, transport, and deposition, is one of the important dynamic land surface processes. The magnetic tracer method is a useful method for studying soil erosion processes. In this study, five types of magnetic tracers were made with fine soil, fly ash, cement, bentonite, and magnetic powder (reduced iron powder) using the method of disk granulation. The tracers were uniformly mixed with soil and tested in the laboratory using simulated rainfall and inflow experiments to simulate the interrill and rill components of soil erosion, in order to select one or more tracers which could be used to study detachment and deposition by the erosive forces of raindrops and surface flow of water on a slope. The results showed that the five types of magnetic tracers with high magnetic susceptibility and a wide range of sizes had a range of 0.99-1.29 g cm?3 in bulk density. In the interrill and rill experiments, the tracers FC1 and FC2 which consisted of ?y ash and cement at ratios of 1:1 and 2:1, respectively, were transported in phase with soil particles since the magnetic susceptibility of sediment approximated that of the soil which was uneroded and the slopes of the regression equations between the detachment of sediment and magnetic tracers FC1 and FC2 were very close to the expected value of 20, which was the original soil/tracer ratio. The detachment and deposition on slopes could be accurately reffected by the magnetic susceptibility differences. The change in magnetic susceptibility depended on whether deposition or detachment occurred. However, the tracer FS which consisted of fine soil and the tracers FB1 and FB2 which consisted of fly ash and bentonite at ratios of 1:1 and 2:1, respectively, were all unsuitable for soil erosion study since there was no consistent relationship between sediment and tracer detachment for increasing amounts of runo?. Therefore, the tracers FC1 and FC2 could be used to study soil erosion by water. Soil erosion, which includes soil detachment, transport, and deposition, is one of the important dynamic land surface processes. In this study, five types of magnetic tracers were made with fine soil, fly ash, cement, bentonite, and magnetic powder (reduced iron powder) using the method of disk granulation. The tracers were mixed with soil and tested in the laboratory using simulated rainfall and inflow experiments to simulate the interrill and rill components of soil erosion, in order to select one or more tracers which could be used to study detachment and deposition by the erosive forces of raindrops and surface flow of water on a slope. The results showed that the five types of magnetic traces with high magnetic susceptibility and a wide range of sizes had a range of 0.99-1.29 g cm 3 in bulk density. In the interrill and rill experiments, the tracers FC1 and FC2 which consisted of? y ash a nd cement at ratios of 1: 1 and 2: 1, respectively, were transported in phase with soil particles since the magnetic susceptibility of sediment approximated that of the soil which was uneroded and the slopes of the regression equations between the detachment of sediment and magnetic tracers FC1 and FC2 were very close to the expected value of 20, which was the original soil / tracer ratio. The detachment and deposition on slopes could be accurately reffected by the magnetic susceptibility differences. The change in magnetic susceptibility depended on whether deposition or detachment occurred. However, the tracer FS which consisted of fine soil and the tracers FB1 and FB2 which consisted of fly ash and bentonite at ratios of 1: 1 and 2: 1, respectively, were all unsuitable for soil erosion study since there was no consistent relationship between sediment and tracer detachment for increasing amounts of runo ?. Thus, the tracers FC1 and FC2 could be used to study soil erosion by water.