Abstract: Hyperspectral imaging technique was investigated for the detection of chilling injury in ‘Qinguang 2'nectarine during the early postharvest storage period. At the beginning, extractable juice content was measured during the storage to find the occurrance time point of early chilling injury. Then, hyperspectral imaging system based on reflectance mode was established to acquire chilled and non-chilled nectarine images during the spectral region of 400¹000nm, as well as to extract nectarine spectral properties by ROI (region of interest) analysis. Dimension reductions were implemented on hyperspectral reflectance data based on Independent Components Analysis (ICA) , for the chilled nectarines at early stage and non-chilled nectarines respectively.The optimal wavelength selected by ICA were 672 nm for chilled ones and 656, 678 nm and 700 nm for non-chilled ones. Next, spectral average of each characteristic band was chosen as the input of the Fisher discrimination model, an average classification accuracy of 98.9% was achieved to distinguish between normal and early injured nectarines. This research demonstrated that the hyperspectral imaging technique was feasible for the detection of chilled nectarine at the early stage.