In order to achieve the rapid identification of buckwheat varieties, and avoid buckwheat varieties mixtures, eight buckwheat varieties from different origins were identified by principal component analysis and support vector machines based on near-infrared spectroscopy. First, the buckwheat spectral information of the 120 samples have been collected using FieldSpec 3 spectrometer, and preprocessing through smooth + Multiplicative Scatter Correction (+MSC), a total of 120 sets were divided into 80 training sets and 40 prediction sets. After the principal component analysis, based on the binary tree support vector machine theory, the spectral information identification model of buckwheat varieties have been established and verified by LIBSVM package in MATLAB software. The results showed that the classification accuracy rate averaged 92.5% for eight different kinds of buckwheat by using near-infrared spectroscopy combined with principal component analysis and support vector machine. A new method for buckwheat varieties identification has been provided.