This paper focuses on investigating the volatility on the Romanian stock market by employing the joint symmetric and asymmetric ARMA - GARCH models on four of Bucharest Stock Exchange’ own indices which reflect only the evolution of market prices: Bucharest Exchange Trading Index (BET), Bucharest Exchange Trading Extended Index (BET-XT), Bucharest Exchange Trading – Investment Funds (BET-FI) and Bucharest Exchange Trading Energy & Related (BET-NG). We estimated ARCH, GARCH, EGARCH and GJR-GARCH models using the maximum likelihood method under the assumption of Gaussian distributed innovation terms. The empirical results show that in three cases out of four the volatility reacted asymmetrically to the good and bad news. The predominant model turned out to be EGARCH model and this is because it does not require any constraint on the parameters since the positivity of the conditional variance is automatically satisfied. JEL Classification: C22, C52, C55, C58