Abstract: In order to determine whether the measurement precision and accuracy of the handheld X-ray fluorescence spectrometer (XRF) can meet the requirements of rapid detection of soil environment, in the context of one soil pollution survey project in a certain plot of Jining City, Shandong Province, the concentrations of arsenic (As), copper (Cu), lead (Pb), zinc (Zn), and nickel (Ni) were detected by XRF rapid detection and laboratory methods. The samples were collected from shallow soil layers (0.2-0.5 m) and deep soil layers (1.5-2.0 m). The results indicated that for shallow soil layers, XRF measurements of heavy metal As consistently exceeded laboratory data, while XRF measurements of heavy metals Cu, Pb, Zn, and Ni were lower within an error of ±30% compared to laboratory measurements. The relative deviations (RD) and overall relative deviation standard deviations of heavy metal elements in shallow soil layers (RD: −33.82%-23.53%, standard deviation: 6.79-19.52) were generally smaller than those in deep soil layers (RD: −30.26%-98.36%, standard deviation: 9.53-49.77), and the dispersion degree of relative deviation in shallow soil layers was smaller than that of deep soil layers. Moreover, the determination coefficients (R²) for heavy metals in shallow soil layers (R²: 0.7762-0.9549) were consistently higher than those in deep soil layers (R²: 0.7762-0.9549), indicating a stronger correlation. When XRF was employed for in-field detection of heavy metals in soil samples, it was recommended to avoid large particles in the samples, compact and flatten the testing surface and, for samples with high moisture content, carry out appropriate drying to minimize detection errors. Despite the inherent limitations of handheld XRF detection, the method demonstrates acceptable accuracy and reliability in reflecting the concentration range of soil elements in soil pollution survey projects, meeting the rapid detection requirements of such projects.