Due to the complexity of geological conditions, the geological and hydrogeological data available to the people arelimited,which leads to the uncertainty in the understanding of hydrogeological conditions. Among them, the heterogeneity of aquifersis the most significant one, which would have a significant impact on the assessment of groundwater vulnerability. This paperconsidered the heterogeneity of aquifers and proposed a groundwater vulnerability assessment method with non -stationary random space spatial correlation. Taking the central area of Jiangning district in Nanjing as an example, a modified successive randomadditional method (SRA) was used to generate a random field with fractional Levy motion statistics for the logarithm of hydraulicconductivity (lnK),and the possible non -homogeneous spatial distribution of the hydraulic conductivity was simulated,and theDRASTIC method was used to evaluate groundwater vulnerability.The results show that the degree of change in the hydraulicconductivity field generated by SRA is more severe than that of the conventional ordinary Kriging method,and it is more in linewith the characteristics of complex distributed non-stationary random fields. The groundwater vulnerability assessment established onthis basis is more in line with objective facts,which enriches and develops theories and methods in using stochastic theory to solvegroundwater environmental problems.