The aim of the present study was to investigate the feasibility of tracing the geographical origin of beef with FT-NIR spectroscopy,set up the model for identifying the beef geographical origin,and validate the rate of discrimination.Fifty eight de-fatted beef samples from Jilin,Guizhou,Ningxia,and Hebei in China were dried and milled.Based on the NIR spectra of the pre-processing beef,the samples were subjected to principal component analysis (PCA),cluster analysis (CA),discriminant analysis (DA),and a qualitative model was established to do discriminant analysis and validated.The results showed that there were some differences in NIR spectra from different geographical origins,and the element contents in the beef samples were different from different locations.According to the Euclidean distance of NIR spectra,the geographical origin can be identified by cluster analysis.The distance of spectra is the shortest between the samples from Guizhou and Hebei province.Meanwhile there is a certain crossover.This model calibrated by 40 samples was used to predict the varieties of 18 unknown beef samples.The recognition rate of 100% was achieved.So applying FT-NIR fingerprint spectroscopy to trace geographical origin of beef is accurate,rapid and low-cost.