In this study, five quality indicators, i.e. moisture, soluble solid content, total sugar content, titratable acid content and hardness, of apple chips during processing period were analyzed by analytic hierarchy process, and quality scores was obtained. Near infrared spectroscopy (NIRS) combined with partial least-squares discriminant analysis (PLS-DA) was applied for the prediction of quality scores of apple chips. The results showed that the correlation coefficients between predicted category variable of calibration and the measured category variable were 0.84, 0.63 and 0.89, and the root mean square error of cross validation were 0.26, 0.34 and 0.22 for each quality category, respectively. The discrimination accuracy for this model were found 83.33%, 80% and 93.33%. Thus, this suggested that NIRS combined with PLS-DA method was a potential way to assess quality of apple chips processing period.