Abstract: Stock exchange index plays an important role in the financial structure of a country and control the up and own movement of price structures of various indexes for business decisions. It is important for business in the stock markets to have an idea about future stock price development. The main objective of this study was to compare the different classification methods for predicting the movement direction of Saudi stock exchange index-TASI (Tadawul All Share Index]. The study applied four classification methods namely Linear Discriminant Analysis (LDA), Logistic Regression (LR), Support Vector Machines (SVM) and Least Squares Support Vector Machines (LS-SVM) as classification methods to predict the movement direction (up or down) of Saudi Stock Exchange Index (TASI) called as Tadawul All Share Index from Aug 2006 to April 2013. The study included several publicly available financial data such as Dow Jones Index, closing price, opening price, SABIC trading volume, crude oil price and TASI as the independent variables. The study chose 5-fold cross validation out-of-sample hit rates to compare the performance. For all the methods, except LS-SVM, hit rates ranged between 61 and 63 and only slightly lower for a reduced set of independent variables . With a data set from April 2009 to April 2013, the hit rates were about 2% higher than the normal rates. Overall, the LR, LDA and SVMs performed well for the price index data used in the study.