This paper examines the role of exchange rates in determining the short-and-long run trade balance behavior for Colombia. Conventional wisdom says that a nominal devaluation improves the trade balance. This conjecture is rooted in the Bickerdike-Robinson-Metzler (BRM) and Marshall-Lerner (ML) conditions. Empirically, the evidence for both developed and developing countries has been inconsistent in either rejecting or supporting the BRM or ML conditions. This paper tests these two conditions. It uses a regression model formulation which includes income and money so that the monetary and absorption approaches to the balance of payments are also examined. The econometric procedure used is the Johansen and Juselius' approach to estimation of multivariate cointegration systems. The main result is that exchange rates do play a role in determining the short-and-long-run behavior of the Colombian trade balance. Moreover, devaluation improves the trade balance, which is consistent whit the BRM or ML conditions. The results show also that the long-run effect of an exchange rate devaluation on the trade balance is enhanced if accompanied by reduction in the money stock and (or an increase in income. The findings whit respect to income and money variables did not uniformly reject or accept hypotheses from the absorption or monetary approaches either for the short run or the long run.