Abstract: In order to analyze the color difference of cigarette packaging material,the near-infrared spectrums of samples of white paperboard coated with inks of different colors were collected and pre-processed via second differential,Karl Norris derivative smoothing and multiplicative signal correction,then a color difference analysis model was established by incorporating Gram-Schmidt orthogonalization.The values of similarity between standard samples and tested samples were calculated by the model,and the calculated values were compared with the results distinguished by colorimeter and visual judgment.The results showed that:1) When the color density of sample was lower than 5% or higher than 20%,the three methods gave identical results.2) When the color density ranged from 5% to 20%,the results obtained by NIR were more accurate than those by colorimeter.3) The color differences of samples from 28 batches were distinguished by NIR,then verified via principal component analysis and Hotelling T² test,the results indicated that it was feasible to analyze the color difference of cigarette packaging material from different batches with NIR spectroscopy.