Abstract: For dense non-aqueous phase liquids (DNAPL) sites, the traditional geological drilling methods are limited by the cost and the number of samples. It is often unable to accurately estimate the residual amount of DNAPL in aquifers. The partitioning inter-well tracer tests (PITT) has low cost and will not cause damage to the source areas, so it can be used to estimate the residual amount of DNAPL. However, the accuracy of this method has not been verified.Based on the numerical simulation method, the effects of various factors, such as tracer injection rate, tracer type and source zone architectures, on the accuracy of estimating the residual amount of DNAPL in PITT are discussed.The results of numerical examples show that the error of PITT in estimating the residual amount of DNAPL is within 15%. Reducing the injection rate of the tracer makes the partitioning tracer more fully interact with NAPL phase and water phase, which can improve the accuracy of PITT estimation of DNAPL residue.After the tracer injection rate decreases from 500 m3/d to 250 m3/d in the experiment, the average accuracy of various scenarios increases from 91.68% to 93%.The selection of tracer with low partitioning coefficient is beneficial to improve the recovery rate of tracer, so as to improve the accuracy of estimation. In the experiment, after the tracer changes from 2,2-dimethyl-3-pentanol to hexanol, the average accuracy increased from 87.83% to 96.85%. Complex source zone architectures are more likely to cause tracer bypass, which makes the error of estimating DNAPL residual larger. In the experiment, after the aquifer changes from homogeneous to heterogeneous, the average accuracy changes from 93.03% to 91.65%. Before using PITT to carry out site investigation, it is advisable to select appropriate tracer and tracer injection rate by combining with numerical simulation method to improve the precision of characterization of contaminated sites.