Analysis of groundwater quality in the alluvial aquifer of the lower Soummam Valley, North-East of Algeria, was realised through the application of multivariate statistical methods: hierarchical cluster analysis (HCA) in Q and R modes, factorial correspondence analysis (FCA), and principal component analysis (PCA), to hydrochemical data from 51 groundwater samples, collected from 17 boreholes during periods of June, September 2016 and March 2017. The objectives of this approach are to characterise the water quality and to know the factors which govern its evolution by processes controlling its chemical composition. The Piper diagram shows two hydrochemical facies: calcium chloride and sodium bicarbonate. Statistical techniques HCA, PCA, and FCA reveal two groups of waters: the first (EC, Ca2+, Mg2+, Cl–, SO42– and NO3–) of evaporitic origin linked to the dissolution processes of limestone rocks, leaching of saliferous soils and anthropogenic processes, namely contamination wastewater and agricultural activity, as well marine intrusion; and the second group (Na+, K+, and HCO3–) of carbonated origin influenced by the dissolution of carbonate formations and the exchange of bases. The thermodynamic study has shown that all groundwater is undersaturated with respect to evaporitic minerals. On the other hand, it is supersaturated with respect to carbonate minerals, except for water from boreholes F9, F14, and F16, which possibly comes down to the lack of dissolution and arrival of these minerals. The results of this study clearly demonstrate the utility of multivariate statistical methods in the analysis of groundwater quality.