To overcome the low efficiency of least squares twin support vector machine (LST SVM) in classifying,a new method called sample reduction LSTSVM (SR-LSTSVM) is prop osed.The method greatly reduces the training samples in kernel space by calculating distance,but the ability of LSTSVM in classifying is not influenced.Meanwhile,to make full use of the spa tial information, principal component analysis (PCA) and two-dimensional Gabor filter are adopted to get texture features of the pixel,and the hyperspectral images are classified by using the t exture and spectral information.Experimental results show that the spatial-spectral information-b ased SR-LSTSVM classification algorithm can not only improve the classification effectiveness but also increase the overall classification accuracy.