This paper presents a novel approach using hyperspectral imaging techn ology for fast and nondestructive detection of apple sugar contents.The optimization of partial least square (PLS) model focuses on a threshold in term of th e PRESS derivative (the sum of squared error) in order to ignore the other inefficient f actors.This novel approach is able to effectively improve the prediction performance and reduce co mputational complexity at the same time.A comparison of three models is conducted,includin g full spectra partial least square model (FS-PLS),genetic algorithm associated with partial least square model (GA-PLS) and successive projections algorithm associated with partial least squ are model (SPA-PLS).The numerical results show that the FS-PLS mo del is the best, which is able to achieve the root mean square error of prediction of 0.614and t he correlation coefficient of prediction of 0.841.The improvement can meet the r equirement in terms of accuracy of fruit inner quality using hyperspectral imaging technology. Meanwhile,the visualized difference of the sugar content is implemented by band mathematical t ool with software ENVI 4.7.