The serious problem of heavy metal compound pollution existing in the lead-zinc mining area, using hyperspectral remote sensing technology to extract the heavy metal absorption characteristic wave spectrum in the soil, to realize the rapid monitoring of the heavy metal pollution of the mining area. We collected 100 soil samples in the field taking Gangdese Lane multi-lead-zinc ore collection area in Tibet as an example. The content analysis and hyperspectral data determination of the sample were carried out in the laboratory. Then the spectral data were subjected to the spectral transformations of first derivative (FD), second derivative (SD), reciprocal logarithmic transformation(AT), first derivative of , reciprocal logarithmic (AFD), second derivative of , reciprocal logarithmic (ASD), to analyze the correlation between the measured content of Cd, Pb, As and Hg and the soil spectra. After selecting the corresponding characteristic bands, we established four inversion models based on the characteristic bands, namely, multiple stepwise regression (SMLR), support vector machine (SVM), artificial neural network (ANN), and random forest (RF) combined with the coefficient of determination (R2) and the root-mean-square error (RMSE) to evaluate the model accuracy. Then the best combination of spectral transformation and inversion model were explored. The results show that: (1) the data dimensionality reduction effect of different spectral transformations is different from the screening interval of the characteristic bands, and Among the five transformations, AT, SD and ASD screening were better than FD and AFD. (2) Comparing the different inversion models, the RF model has the best inversion effect, and its applicability and inversion accuracy are better than SMLR, ANN and SVM.(3) The best inversion model for As is the AT-RF model, the best inversion model for Cd is the SD-RF model, the best inversion model for Pb is the ASD-RF model, and the best inversion model for Hg is the ASD-SMLR model. This study can provide technical support and reference for the monitoring of soil heavy metal content and pollution evaluation of large-scale alpine and high-altitude lead-zinc mines. (4) The predicted concentrations of Cd, Pb, As and Hg were 124.2, 89.8, 0.70 and 1.24 times higher than the background values of Tibetan soil, indicating that Cd and Pb were the main pollution factors in the soil in the mining area, and were accompanied by the combined pollution of As and Hg.