It is urgent to think up a quick and precise method in the discrimination of the famous tea. A new discriminate method of tea is proposed. In this study, Biluochun tea serves as the target and the model for discriminating the authenticity of tea is built up using near-infrared spectroscopy combined with the pattern recognition of support vector machine (SVM). The experimental result shows that in the spectra region between 6500 cm-1 and 5300 cm-1, the best model is built by the standard normal variate (SNV) preprocessing, when 11 principal components are selected and the radial basis function (RBF) is used as the kernel function. The discriminate rate of the model for 138 samples in training set is 93.48%, and 84.44% for 90 samples in predicting set. The research shows that it is feasible to apply near-infrared spectroscopy to discriminate the authenticity of the famous tea based on SVM.