Abstract: In order to research the technical method and workflow of rapid investigation of soil heavy metal content and distribution characteristics in large-scale and large-area by airborne hyperspectral remote sensing, taking Guanhe area in northern Jiangsu Province as an example, the domestic hyperspectral sensor and the independent integrated airborne hyperspectral remote sensing system were used to obtain the image spectral data. By means of data preprocessing, CARS characteristic band selection and BP neural network modeling inversion, a workflow from data acquisition to data analysis and evaluation was established, the content of heavy metals in soil was estimated and its spatial distribution characteristics were displayed. Compared with the traditional geochemical survey data, the results show that the hyperspectral inversion results are in good agreement with the geochemical results, which verifies the reliability of the airborne hyperspectral inversion technology. In the aspect of heavy metal pollution investigation and evaluation of regional farmland soil, airborne hyperspectral remote sensing can quickly and timely obtain soil pollution and related information with good practicability and economical efficiency.