Abstract. This study investigates the weak form efficiency of Efficient Market Hypothesis (EMH) employing Autocorrelation test, Runs test and Unit Root tests, and the nature of volatility characteristics of stock returns applying GARCH family models in Bangladesh stock market using daily all share price index return data of Dhaka Stock Exchange (DSE) from 02 January 1993 to 27 January 2013. This studyalso examines the semi-strong form of the EMH of DSE based on macroeconomic variable version of the Arbitrage Pricing Theory (APT) applying Cointegration tests, Vector Error Correction Model (VECM) and Granger causality tests, and the volatility of the DSE returns in response to the volatility of the macroeconomic variables employing GARCH family models using monthly data from January 2001 to December 2012.In addition, the short run and long run relationships between macroeconomic variables and aggregate stock prices in Bangladesh have also been determined. Employing both nonparametric tests (Runs test and Phillips-Perron test) and parametric tests (Autocorrelation test and Augmented Dickey-fuller test), this study finds that the DSE of Bangladesh is not weak form efficient. Taking the outcome of VAR models into account, it is found that all selected macroeconomic variables do significantly explain the stock prices of the Bangladesh stock market. As a consequence, it may be concluded that the Bangladesh stock market is not efficient in the semi-strong form of EMH. Results of the estimated MA(1 ) -GARCH(1,1) and MA(1 )- EGARCH(1,1 ) models reveal that stock market returns of Bangladesh exhibit leptokurtosis, volatility clustering and leverage effect. Results of six GARCH-S models indicate that thevolatility of DSE return is significantly influenced by the volatility of macroeconomic variables, such as, exchange rate, broad money supplyandstock returns of India. Keywords. Efficient market hypothesis, Stock prices, Vector error correction model, GARCH family models, Volatility. JEL. C58, E44, F36, G10, G14.