Abstract: Models for maturity discrimination and polyphenol content determination of Aronia melanocarpa (AM) were established to nondestructively determinethe maturityand polyphenol content of AM based on the hyperspectral imaging (HSI) technology. The HSI technology was adopted to collect the image information of Fukangyuan No. 1 AM in different maturity levels, and the Folin-Ciocalteu colorimetric method was employed to determine the polyphenol content. The Monte Carlo method was used to eliminate the outliers. The original image information was preprocessed with the following procedures: Moving average, median filter, normalize, baseline calibration, multiple scattering correction, detrending, and standard normal variate. Meanwhile, the sample set partitioning based on joint x-y distance method was applied to divide the samples. The competitive adaptive reweighting sampling and uninformative variable elimination method were selected to extract the feature wavelengths, based on which the partial least squares model (PLS) and support vector machine (SVM) model were established and compared. The results showed that the UVE-SVM model after multiple scattering correction showed the best performance among all established discriminant models, with the comprehensive recognition rate of 94.62%, R_c² of 0.9712, and accuracy of 100%. The CARS-SVM model after median filter exhibited the best efficiency in detection of polyphenol content, with the R_c² of 0.8331. In addition, this work proved that the polyphenol content in AM was visual. This work provided a theoretical basis for the application of HSI technology in berry industry.