In this study, an improved DRASTIC model, including the DRASTIC-LTPD model and the AHP-DRASTIC-LTPD model, with the addition of four extra evaluation factors, including land use type (L), aquifer thickness (T’), aquifer potential (P) and pollution source distance (D’), was constructed and compared to assess the groundwater vulnerability around farmland area in Shuangsheng Industrial Park, Sichuan Province, China. From the vulnerability grading charts of the traditional DRASTIC model, the improved DRASTIC-LTPD model and the AHP-DRASTIC-LTPD model, it showed that the vulnerability presented a lower level in the western and eastern farmland areas, whereas a higher level was in the central industrial park area. This result was consistent with the actual situation where groundwater recharge by rivers, regional land use, and human activities were more active in the middle in these areas. Nevertheless, the area at the same level of vulnerability varied greatly from model to model. The vulnerability index V-level region ratio calculated by the AHP-DRASTIC-LTPD model was 0, indicating that the distribution of vulnerability was smoother without the appearance of extremely good or poor conditions. From the present study, it was revealed that the AHP-DRASTIC-LTPD model could effectively reflect the impact of human activities and dilution on groundwater vulnerability. The adopted AHP method was also of high accuracy to empower the evaluation index leading to a more reliable evaluation results of regional groundwater vulnerability in comparison with the other two models. Therefore, this research could be employed as a reference for the evaluation of groundwater pollution around other similar unplanned industrial parks.