Assessing groundwater sensitivity in areas contaminated by agri-cultural pollution is a complex task for groundwater resource managers. The Ghriss plain (north-western Algeria) is the basis of agricultural development based on diversified crops, irrigated from a local water table of Plio-Quaternary age, considered to be one of the country’s most important reservoirs. In recent years, the overuse of nitrogenous fertilizers and various phytosanitary products, combined with wastewater discharges, has led to the appearance of traces of chemical elements exceeding drinking wa-ter standards, and to a worrying risk of water quality deteriora-tion. This study aims to measure the different categories of vul-nerability to migration of chemical inputs through the unsaturated zone, to preserve groundwater contamination. The approach con-sists of applying two universal methods: the GOD model and the SI model (Susceptibility Index). The mapping of the two models shows that GOD represents a low concentration of measured ni-trate ions, while SI attributes a high pollution dispersion to the unsaturated zone. From this comparison, the GOD method indi-cates the presence of four classes of vulnerability: high 1%; medi-um 70%; low 28% and very low 1%. On the other hand, the SI comparison shows two classes of vulnerability to pollution: very high to high 72% to 23%; medium to very low 2% to 3%. From this comparative study, we deduce that SI gives us a true picture of the pollution phenomena likely to occur, taking into account the nature of the land use, hence its interest in a more accurate esti-mation of unsaturated zones' vulnerability to groundwater pollu-tion by nitrates.