Soil erosion is known as one of the most significant environmental problems throughout the world. Several studies showed that soil erosion rate in Iran is too much and annually causes huge irreparable threats such as filling reservoirs of dams, floods, soil and water pollution and declining soil productivity. Therefore, sediment source determination, for identifying of contribution of various sources of sediment production within the watersheds is essential for prioritizing of managements practices. The sediment source determination is the determination of the contribution of each sediment source based on the results of various tracers such chemical, physical and magnetic properties. Despite the several studies using chemical and physical tracers, little attempt has been done using magnetic susceptibility, especially in Iran. This study was conducted to determine the relative contribution of various sources (land use, geology, erosion, sub-basins, land use-erosion and geology-erosion) in the production of sediment by fingerprinting technique using Bayesian Uncertainty Model, in Sanandaj basin. For this purpose, 83 samples of surface erosion areas (with specific geological settings), 58 samples of rill and channel erosion areas (with specific geological setting), 20 samples from the five sub-basins of the region and 8 samples of sediment production from the basin outflow were collected. To determine of sediment contributions, 15 tracers including low-frequency magnetic susceptibility (? lf ), high-frequency magnetic susceptibility (? hf ), dependent frequency (? fd ), the geochemical elements concentrations of K, Ca, Mg, Cu, Zn, Fe, Mn, Cr and Co, total nitrogen (N), organic carbon (OC), and available phosphorus (P) were selected. Kruskal-Wallis test was employed to determine the ability of selected tracers to distinguish at least two sources, then using the discriminant analysis (DA), the optimal combination of tracers was determined each source. In the next step, using the Bayesian uncertainty model, the contribution of each source to sediment yield was determined. The results showed that for land use contribution, pasture (41.9 %), for the geology source, PE unit (74.2 %), for the erosion type sources, sub-surface erosion (62.7 %) and for the sub-basin contributions, Todarsamadi sub-basin (55.9 percent) had the largest contribution in sediment yield. The results of sediment sources were in consistent with the results of the evaluation of organic carbon storage and nitrogen in the studied region. The integration of chemical tracers with magnetic susceptibility data increased the determination of sediment’s certainty in the sub-basin source and type of erosion. As a result, it is concluded that the magnetic susceptibility technique can be effectively used in determination of sediment sources studies. The results of this study could be applied to select the appropriate method for erosion control in the studied region and similar areas. Keywords: Soil erosion, Determination of sediment, Soil fingerprinting, Magnetic susceptibility, Bayesian uncertainty model.