A near infrared(NIR) spectral database for pharmaceutical powders of Chinese medicine oral solid dosage forms was established,and the application of NIR spectral data in material classification and material properties prediction was studied by pattern recognition method.NIR diffuse reflectance spectra for 149 batches of pharmaceutical powders were rapidly measured with portable near infrared spectrometer,and the NIR spectral data were input into the intelligent traditional Chinese medicine(iTCM) database.Principal component analysis(PCA) method was used to explore the classification ability of NIR spectral data for materials with known structures.Partial least squares(PLS) method was used to study the prediction ability of NIR spectra on the material properties of pharmaceutical powders and tablet properties in direct compression.The NIR spectra pretreated by standard normal variable transformation(SNV)+Savitzky-Golay(SG) smoothing+first derivative exhibited a good discrimination ability for five kinds of excipients,i.e.microcrystalline cellulose,lactose,ethyl cellulose,polyvinylpolypyrrolidone and hydroxypropyl methylcellulose.NIR spectra had a strong correlation with particle size,density and hygroscopicity.NIR spectral information could be used as a supplement for the material properties,which could improve the performance of the prediction model for tablet properties in direct compression.Meanwhile,NIR spectral data were also a kind of supplement for the material properties data of iTCM database.The combination of material properties and NIR spectral data could be used to comprehensively characterize the properties of pharmaceutical excipients.