The exchange rate strongly affects the cross-borders economic transactions. Investment, trade, migration, tourism and more are all intensely influenced by the fluctuation of exchange rate. This study re-examine the determinants of nominal exchange rate. Our main contribution is to find an appropriate model of exchange rate through general to specific approach. Then the empirical confirmation of the long run relationship of nominal exchange rate with independent variables is found through this study employing annual data set for the period 1980 to 2015. We also compare Bayesian techniques and classical technique on the basis of out of sample forecasting. Autoregressive Distributed Lag Model (ARDL) Approach to cointegration or Bound Cointegration Testing Approach is used to find the long run association. The results shows that Prices and terms of trade as significant influence on exchange rate in long run, using classical as well as Bayesian approach. It is found out that Bayesian approach is preferred while using G-prior as an estimation technique as compared to classical approach. Supervisor:- Dr. Atiq ur Rehman