Water quality characterization is important, and often, the first step in watershed water quality management. The characterization activity centers on the quantification of water quality parameters relevant to the monitoring objective and the analyses of the impact of the parameters on the state of water body. This study characterizes the general water quality of a network of agricultural rivers in Southern Ehime, Japan and conducted Factor Analysis (FA) and Cluster Analysis (CA) to determine the causal factors and spatial variation of pollution. The Water Quality Index (WQI), based on nine water quality parameters, was computed and optimized-determining the parameters that could best predict the overall water quality. Results show that the river sampling sites have good overall water quality and that physico-chemical parameters pH, DO, and BOD_5 best correlated and predicted WQI. Based on the FA, the primary factors that influence pollution in the sites are the solids or sediment transport (turbidity and sediment), inorganic pollution (NO_3-N, PO_4-P) and trophicity (total N, total P, fecal bacteria). CA, on the other hand, somehow provides information on the effect of the tributaries to the receiving rivers.