Redox couples approach and multivariate statistical techniques, including factor analysis, cluster analysis and discriminant analysis, were applied to evaluate and to interpret the complex groundwater quality in the blackfoot disease endemic area, Taiwan. Most groundwater samples were characterized as Na–Ca–HCO 3 with HCO 3 − as the dominant anion. Total arsenic (As) concentration, predominantly as As 3+, ranged from <1.0 to 562.7 μg/L. The patterns of measured reducing potential were consistent with those values calculated from As couple, revealing the in situ environment enhanced the accumulation of As concentration in the groundwater. Factor analysis proposed a four-factor model, comprising salination, reductive dissolution of Fe/Mn oxyhydroxides, As reduction and chemical potential factor, and explained 89.94% of total variance in groundwater. Furthermore, two factors, reductive dissolution of Fe/Mn oxyhydroxides and As reduction, suggested that the decoupled reductive processes accounted for high As concentration in this area. Cluster analysis was adopted to spatially categorize the sampled wells into three main clusters and characterized by the factor scores of the four-factor model. Two-parameter (pH and Eh) model derived from discriminant analysis can be used for preliminary assessment to determine whether the As concentration exceeds 10 μg/L with simple field measurements in this area .