Abstract: Inversion modeling of heavy metal nickel in soils of typical iron mine area in Beijing was carried out with hyperspectral data, in order to explore the feasibility of hyperspectral remote sensing technology in rapid monitoring of soil heavy metal pollution. The spectral reflectance data of soil samples in the study area were obtained by portable spectrometer, and the original spectral data were mathematically transformed. Optimum feature bands were extracted with Stepwise regression method. The inversion model of soil nickel content was established with Multiple Linear Regression (SLR) and Partial Least Squares Regression (PLSR). The multiple linear regression model based on spectral second derivative (SD-MLR) had the highest stability and accuracy (R² = 0.842, RMSE = 4.474), which was the best inverse of the nickel content in the soil of the study area. Spectral mathematical transformation can effectively improve the correlation between spectral data and element content. The following relationship exists in the modeling predictive power and accuracy of spectrum reflectance under different transformations: second derivative > absorbance transformation first derivative > first derivative > absorbance transformation > continuum removed > original spectra.The SD-MLR model could well predict the nickel content in the area and provide a technical reference for rapid monitoring of soil heavy metal pollution.