In the current groundwater contamination investigation study, sampling point allocation methods generally rely on modeling data analysis, which undoubtedly leads to increased sampling costs. Based on this lack of sampling data, the AHP-entropy weight method is improved and applied to construct a distribution model of groundwater pollution sampling points in a plant area, reduce the sampling scale and establish a new and optimised samplimg scheme. A chemical plant in Shenyang is selected for method validation and spatial interpolation analysis, and the following conclusions are drawn. (1) The new scheme can identify areas of contaminants and reduce sampling costs. (2) After reducing the sample size, the spatial distribution characteristics of each pollutant can still be clearly distinguished. (3) The interpolation-based predictions of some pollutants were improved, and the accuracy of the predictions of other pollutants was reduced within the permissible range.