A method of visible spectroscopy analysis technology was used to evaluate the visual characteristics on the surface of decorative paper in order to explore the feasibility of visible spectroscopy instead of traditional human’s visual assessment and instrument measurement to classify decorative paper.The classification models were established by visible spectroscopy data from six types of decorative paper using the methods of partial least-squares discriminant analysis（PLS-DA）,and the classification accuracy derived from different spectral pretreatment methods were compared.The results showed that the classification accuracy based on raw spectra was 93%～100%.The accuracy using multiplicative scatter correction（MSC）and standard normalized variate（SNV）pretreated spectra was up to 97%～100%.