Water pollution has become a growing threat to human society and natural ecosystems in recent decades, increasing the need to better understand the spatial and temporal variabilities of pollutants within aquatic systems. This study sampled water quality at 12 sampling sites from October 2006 to August 2008 in the Jinshui River of the South Qinling Mts., China. Multivariate statistical techniques and gridding methods were used to investigate the temporal and spatial variations of water quality and identify the main pollution factors and sources. Two-way analysis of variance (ANOVA) showed that 25 studied water quality variables had significant temporal differences (p<0.01) and spatial variability (p<0.01). Using cluster analysis, the 12 sampling sites were classified into three pollution level groups (no pollution, moderate pollution, and high pollution) based on similarity of water quality variables. Factor analysis determined that 80.4% of the total variance was explained by five factors, that is, salinity, trophicity, organic pollution, oxide-related process, and erosion. The gridding methods illustrated that water quality progressively deteriorated from headwater to downstream areas. The analytical results suggested that the water pollution primarily resulted from domestic wastewater and agricultural runoff, and provided critical information for water resource conservation in mountainous watersheds of the South Qinling Mts., China.