Abstract: The fractional order derivative (FOD), an important branch of mathematics, extends the classical integer order to an arbitrary order, allowing for capturing detailed spectral features. This study established a hyperspectral image-based inversion model for soil heavy metal content using hyperspectral images and chromium (Cr) element from field soil samples. Initially, pixel spectra were extracted based on coordinate information of field soil samples, followed by sample expansion by extracting neighborhood pixel spectra according to the principle of minimum Euclidean distance. Subsequently, the expanded samples were processed using Savitzky-Golay (SG) filtering and multiplicative scatter correction (MSC), followed by FOD spectral transformation. Then, feature band combinations were selected using the competitive adaptive reweighted sampling (CARS) algorithm. Finally, a partial least squares regression (PLSR) model was established based on the selected feature combinations, followed by an evaluation of its accuracy. The results indicate that the expanded samples effectively alleviated the "overfitting" of the inversion model, enhancing inversion accuracy and stability. In case of a FOD order of 1.8, the PLSR inversion model exhibited the highest accuracy on both training and testing sets, demonstrating the strongest adaptability and stability. Specifically, the training set exhibited a coefficient of determination (R²) of 0.896 2 and a ratio of performance to deviation (RPD) of 3.104 4, while the testing set presented a R² of 0.755 6 and a RPD of 2.022 6. Both sets showed the lowest root mean squared error (RMSE) at 16.331 and 17.094 mg/kg, respectively, indicating the model reached an approximate prediction level. Building upon hyperspectral data from domestic satellite ZY-1 02E, the proposed inversion model for soil Cr content in pyrite mining areas contributes to the relevant research on heavy metal content inversion using domestic satellite-based hyperspectral data, providing technical support for the prevention and control of heavy metal pollution in pyrite mining areas.