Abstract: Abstract: In the upstream of water sources, many pollutants are output from key pollution areas of water sources, and thus identifying key pollution areas of water sources is a key to protection and pollution control of rural drinking water source. In this paper, a method to directly identify key pollution areas was proposed and applied to investigate the pollution situation of rural drinking water sources in Anjishan reservoir area in 2011. Residential areas, farmlands, mines, and other land use types were identified as the pollution sources based on image classification techniques of remote sensing. Hydrology analysis tools for ArcGIS software were used to determine catchment of Anjishan basin. Batch Sub watershed Delineation tool combined with the distribution of the main pollution sources was used to divide Anjishan basin into 19 sub watershed, and land use areas for each sub watershed were summarized. In rural areas, Anjishan reservoir area was mainly affected by pollutants from non-point sources. So, farmland, aquaculture sewage, mines and rural life were considered as main pollution sources of this area. For these sources, COD (chemical oxygen demand), NH3-N (ammonia nitrogen), TN (total nitrogen) and TP (total phosphorus) were the main pollution factors. Source strength coefficient of each factor for each source was obtained from national reports and amended with survey data. Exceeds bid index method was adopted to calculate the weights of pollution factors based on their contributions rates, and degree of pollution was also calculated based on the weights. The pollution degree of water source for each sub watershed was calculated through degree of pollution of each pollution source and its corresponding land use area. The pollution degree of each sub watershed was converted to pollution concentration based on annual rainfall. Environmental quality standards for surface water in China are often used to assess the condition of water sources. In the standard, Class III is for secondary area of centralized drinking water, class IV is for industrial water and recreational water without direct contact to human body, and class V is for agricultural and landscape water. According to the weight of each pollution factor, the surface water environment index standard S was amended. By comparing the pollution concentration and the amended standard values, the sub watersheds was classified into 5 grades including normal area (S is 0-1.35), general pollution key area (S is 1.35-2.04), secondary key pollution area (S is 2.04-2.75), and key pollution area (＞2.75). Among the 19 sub watershed, 4 watershed areas were identified as normal area including watershed 8, 14, 16 and 17; 3 general pollution key areas were 9, 10 and 13; the second key pollution areas were 2; and key pollution areas were 1, 3, 4, 5, 6, 7, 11, 12, 15, 18 and 19. The sub watershed 5, 6, 11 and 12 were identified to be polluted seriously. It indicated that the pollution of drinking water sources in Anjishan was mostly from farmland runoff and residential life. The study suggests that the new method is viable for key pollution area identification of rural drinking water sources.