Abstract: An exploration on heavy metal Pb2+ detection method was presented based on Terahertz(THz) Spectroscopy, which is rapid and maybe portable and low cost for future in-field applications. THz is a brand new and safe technology with many unique features. It showed feasibility for heavy metal detections in soil according to former experiments. A series of experiment were carried out in this study. Soil samples with predefined concentration levels of Pb2+ ions were carefully prepared. Pressed-slice method was used for the spectra measurement and sample making parameters were determined. The absorption spectra were collected with a THz spectroscopic system in Capital Normal University. Multivariate statistical methods were studied and used to analyze the collected data and establish predication model for Pb2+ concentrations. Calibration and prediction models were established based on Partial-least-square(PLS) and Interval Partial-least-square(i-PLS) methods using the full THz spectrum and the selected THz wavebands respectively. The results show that the optimized models are able to predict soil heavy metal Pb2+ content with a correlation coefficient of 0.81. The method of using THz technology for soil heavy metal Pb2+ detection is feasible. With the detection results, it can be found the heavy metal Pb2+ pollution spatial distribution information of the soil in-field and get the pollution map. Specific measures shall be taken based on the pollution map to manage the soil to improve the farmland productivity. This study will provide a reference for heavy metals pollution detection solutions for precision farming and sustainable agriculture development.