The objective of this study is to establish the background values of major and trace elements in sediments of the Xiangjiang River by understanding the element composition of the sediments, and discriminating between the natural and anthropogenic sources of heavy metals in the sediments. Thus, the geochemical background values of elements with different sources in the sediments were calculated by using different estimating methods .Methods： Samples for this study were collected from sedimentary cores obtained by drilling into the bed sediments of the river, then the major and trace elements were analysed using XRF and ICP- MS techniques, respectively. Principal component analysis (PCA) and enrichment factor (EF) values were used to identify element sources. For elements from natural sources, this study used reference element (Al2O3) regression analysis, Tukey’s boxplots of local EFetc., to eliminate outlier values and calculate the background values of elements. For elements with input of anthropogenic sources, the cumulative distribution function (CDF) curve was used after the calculations of regression analysis and Tukey’s boxplots of local EF to eliminate the samples that were polluted by human activities.Results：The results suggest that major elements SiO2, TiO2, Al2O3, Fe2O3, and K2O, and trace elements V, Co, Cr, Ba, Sc, Th, U, Sr, Ga, Ge, Rb, Nb, Y, and REE relatively have less variation (Cv< 0.20) of concentrations in the sediments, and they are believed to be homogeneously distributed in the sediments. These elements do not show significant enrichment or depletion features in the sediments (EF< 2.0, except the depletion of Sr). While, major elements MnO, MgO, CaO, Na2O, and P2O5, and heavy metals Cd, Mn, Cu, Pb, and Zn generally have considerably high variation (Cv > 0.25) of concentrations in the sediments, and they were distributed very heterogeneously in the sediments, among which the heavy metals are significantly higher enriched (EF> 2.0) if compared to that of the NASC. Conclusions：The principal component analysis (PCA) and Pearson's correlation analyses suggest that the elements with EFvalue around 1.0 are hosted mostly within the silicate minerals, and they are of natural sources. While those heavy metals that are significantly enriched in the sediments (EF > 2.0) reside mostly within the Fe—Mn oxide minerals, and they may include the proportions that were contributed from the anthropogenic sources.Thus, different estimation methods were applied to calculating of the background values of the elements that have different source features as above. Then the background values of 47 elements in the bed sediments of the Xiangjiang River were established in this study, and the reliability of this background was testified by this study.