We investigate the sources of real exchange rate fluctuations. We do so, first, in the context of a DSGE model that explicitly considers the central bank's preferences. Then we estimate SVAR models, where shocks are identified by sign restrictions derived from the DSGE model. We perform this exercise for twelve countries, nine of which have adopted inflation targeting during the period analyzed. In sharp contrast to the previous evidence in the literature, we find that exchange rate (country risk premium) shocks have become the main drivers of real exchange rate dynamics, while real shocks play a less important role. Evidence from the DSGE model reveals that, as the central bank becomes more averse to inflation movements, and cares less about nominal exchange rate fluctuations, the impact of nominal shocks on the real exchange rate tends to increase, while the impact of real shocks decreases. Our results suggest that the adoption of inflation targeting, along with a floating exchange rate, contributes to a shift in the relative importance of demand and country risk premium shocks in determining the RER.