Abstract: Abstract: Watershed management for soil conservation has great beneficial effects on the Loess Plateau in China, which can be characterized by a very high specific sediment yield (SSY) of >10000 t/km2·a. The major conservation measures include the check-dam construction (channel measures) and terracing, afforestation, pasture establishments (slope measures). In various conservation measures, it is still lacking efficient method to partition their synergistic impacts, as well to reduce SSY. Here, this paper presents a new method to monitor the relative effects of rainfalls and different conservation measures on the SSY change in the Loess Plateau. In the framework of the runoff-sediment yield relationship, the sediment-reduction mechanisms of rainfalls and various conservation measures can be expressed by: SSY=CR, where R denotes the surface runoff and C the mean sediment concentration in surface runoff. The results show at the watershed scales: 1) Slope measures can only reduce SSY by decreasing R as the flows entrain some sediment from upland slopes to gullies; 2) Rainfalls has the same effects as the slope measures because the rainfall cannot reduce the available sediment, where the mass movement is very active on the Loess Plateau; 3) Check dams can reduce SSY by decreasing both R and C. Based on the differences in the sediment-reduction mechanism, the equations were developed to calculate the SSY-reduction effects of rainfalls (ΔSSYP), slope measures (ΔSSYSM,) and check dams (ΔSSYCD). The resultant equations are: 1) ΔSSYP=C0ΔRP, where C0 denotes the C value for the reference period (P0), where soil conservation practices were the minimum, and ΔRP denotes the R change due to the rainfall change; 2) ΔSSYSM=C0ΔRSM, where ΔRSM denotes the R change due to the slope measures; 3) ΔSSYCD=C0ΔRCD+RiΔC, where ΔRCD denotes the R change due to the check dam establishment, Ri denotes the R value for the ith evaluation period (Pi), and ΔC denotes the C difference between the P0 and Pi. The term C0ΔRCD represents the reduced SSY by check dams reducing R, and the term RiΔC represents the reduced SSY by check dams reducing C. The results showed that there was a good agreement with the observed and calculated equations (R2>0.9). The developed equations are different from those in the "hydrology" and the "soil and water conservation" method, where there is no the SSY-reduction effect at the watershed outlet, but the erosion-reduction effect occurs on slopes due to neglecting the variations of the SSY-reduction effects with spatial scales. The proposed method considers the spatial scale variation of the SSY-reduction effects, to obtain the SSY-reduction effects at the watershed outlet. This method can also provide a sound theoretical foundation to monitor the relative contributions of various conservation measures to the SSY change on the Loess Plateau.