Lycium ruthenicum Murr. is a kind of traditional food with abundant nutrition such as polysaccharides, proteins, minerals and anthocyanins. It has a long history used as medicinal and food plants in China, meanwhile it has functions of scavenging free radicals, anti-oxidation, beautifying and regulating the human immune system. Lycium ruthenicum Murr. is mainly distributed in Tibet, Xinjiang, Inner Mongolia, Qinghai and Ningxia and so on. Different kinds Lycium ruthenicum Murr. have different kinds of quality. All of that can be calculated to high altitude, big diurnal amplitude and environmental aspect in different regions. Thereby, with the increase of demand for black Goji berry, there are miscellaneous black Goji berry priced at different price in the market. In order to rapidly and efficiently deter minute geographical origin and categories in Lycium ruthenicum Murr., Near infrared (NIR) and Fourier transform infrared (FTIR) spectroscopy was employed with the help of chemometrics. Five kinds of Lycium ruthenicum Murr. were analyzed. The 175 Lycium ruthenicum Murr. can be classified into 5 groups. Least-squares support vector machine (LS-SVM) was first performed to calibrate the discri mination model to identify the geographical origins and categories of Lycium ruthenicum Murr. LS-SVM model based on the combination of two spectroscopies were superior to those from either FTIR or IR spectra and the recognition rate of LS-SVM reached up to 99.17%, which showed excellent generalization for identification results. Polysaccharide contents were closely related with the quality of Lycium ruthenicum Murr. Synergy interval partial least squares (Si-PLS) was applied to develop the prediction model of polysaccharide contents. The model was optimized by a leave-one-out cross-validation, and its performance was tested according to the root mean square error of the cross validation (RMSECV) and correlation coefficient (R_c) in the calibration set. Experimental results showed that the optimum results of the Si-PLS model were achieved as follow: RMSECV=2.08%, R_c=0.976 9 and root mean square error of prediction (RMSEP)=2.40%, and correlation coefficient (R_t)=0.967 0 in the prediction set. Finally, the robustness of the LS-SVM model obtained was checked with the 15 new samples that did not belong to the calibration set. And, the calibration model obtained during the work was applied and the calibration values were compared with the external validation values. Si-PLS model provided RMSEP and R_t were 0.947 7 and 2.57% in external validation The overall results sufficiently demonstrate that the spectroscopy coupled with chemometrics has the potential to distinguish Lycium ruthenicum Murr.