Heavy metal concentrations of the sediment from an inland river in Suzhou city, northern Anhui Province were determined, and the index of Geo-accumulation (Igeo) was calculated for evaluating the extent of heavy metals contamination. The results showed that the contents of Cr, Cu, Zn, As, and Pb are higher than those of the background values, and the Igeo data indicates that these metals can be subdivided into three kinds, un-pollution (Cr and As), slight pollution (Pb) and moderate pollution (Cu and Zn). Multivariate statistical approaches were applied for source discrimination of heavy metals. The multivariate statistical approaches drew the final conclusion: (1) Cluster I consisted with Cu and Pb is greatly impacted by the traffic emissions, corresponding to the PC1; (2) the high accumulation of Zn and As in Cluster II mainly derived from the pollutant emissions in industry area, representing the PC2; (3) Cluster III (only Cr) is considered to be controlled by the parent materials of soils and shows the character of PC3; (4)The high contamination of Cluster I and II are related to the human activities, the government should pay attention to the environment in the area.