Volatility models cover a wide range of topics in econometrics and Statistics. They arernbroadly divided in to two categories: ARCH type and SV models based on therndetermination of variance at time (t-I) having all information at time t. The objective ofrnthis study is modeling the volati lity of exchange rate of ETB deterministically andrnstochastically using ARCH type and SV models, respectively. ARCH type volati lityrnmodels are considered as deterministic whereas SV models are stochastic. The data seriesrnused for the study is obtained from the NBE from February 2, 2001 to March 15 , 20 13rnand consists of about 3092 dai ly observations. The major currencies selected for the studyrnbased on the availability of data documentation were: EURO, GBP and USD FOREXrnrate per ETB. A variety of time series models such as ARCH (1), ARCH (2), sGARCHrn(1 , 1); EGARCH (1 , I), APARCH (1 , 1) and basic SV model were estimated. The modelsrnused for the analysis were selected based on their performance reflected in the literaturernof econometrics and Statistics. Among the major currencies selected for the study; arnEURO FOREX rate failed to pass the ARCH effect test and model estimation was donernonly for GBP and USD FOREX rate. The results of the study indicate that sGARCH (I,rnI) outperforms other ARCH type time series models for modeling the volatility ofrnexchange rate of ETB per major currencies (GBP and USD). Furthermore, SV model wasrnmodeled alone without any comparison for both currencies return. SV model fulfilled allrnthe basic assllmptions and can be a candidate for fitting the volatility of GBP and USDrnreturn series per ETB.