In order to develop a fast and efficient method to determine the freshness of bee pollen, visible and near infrared (Vis/NIR) spectroscopy combined with least square support vector machine (LS-SVM) was applied to identify the samples of bee pollen under the different storage temperature. Five storage temperatures including -20, 4, 15, 25 and 40℃ were set. The Camellia pollens stored for 60 days at different set temperatures were investigated. Spectra were acquired by an ASD fieldspec spectrometer. As the pretreatments of the optimal smoothing way, moving average with three segments and multiplication scatter correction (MSC) were applied to reduce the noise of the spectra. After the principle component analysis of the spectra from 400 to 1000 nm, 4 to 20 principal components (PCs) were chosen as the inputs of LS-SVM models. Results show that the r2p of the LS-SVM model with 20 PCs was more than 0.9919 in validation set, and was better than the traditional PLS model. The results also indicated that the PCs could reflect and represent the main characteristics of bee pollens stored at different temperatures, and the LS-SVM model could be used to discriminate the samples of bee pollen under the different storage temperature.