The aim of this thesis is to combine economic theory and empirical analysis in an effort to understand the dynamic effects of real exchange rate determinants, policies and global factors on real exchange rates. This thesis comprises three related essays.The first essay examines the validity of the Balassa-Samuelson hypothesis (BSH). This study introduces a new approach for classifying traded and non-traded industries which allows for country-specific heterogeneity and trade endogeneity, and then uses this classification in the construction of a model that allows for the Balassa-Samuelson effect. We find that in developed countries, productivity growth in traded sectors leads to a real depreciation, inconsistent with the BSH; however, higher economic growth will be followed by a real appreciation. The results of developing countries support the BSH, although persistence profiles show slow speeds of convergence.The second essay extends the analysis into a general model of real exchange rates. It investigates the impact of trade liberalisation, productivity growth, monetary policy and government consumption on real exchange rates in four panels of countries consisting of European, non-European developed, Asian developing and non-Asian developing countries. The analysis is based on a panel structural vector error correction model augmented with foreign variables, and a Bayesian approach is used to implement sign restrictions with a penalty function for undertaking impulse response analysis. We find that trade liberalisation generates depreciation and higher government consumption causes persistent appreciation. A contractionary monetary policy shock has only short-run impact on real exchange rates, corresponding to the long-run neutrality of monetary policy. Traded-sector productivity gains cause an impact appreciation in Asian developing countries and lead to persistent appreciation in non-Asian developing countries, whereas the shocks induce long-run depreciation in developed countries, in line with the results in the first essay.The third essay combines the four panels of countries into a Global Vector Autoregressive (GVAR) model to examine how real exchange rates and key macroeconomic variables respond to an oil price shock, a US monetary policy shock and simultaneous shocks to productivity in four large Asian emerging economies. Using a sign restricted impulse response approach, we find that an oil price shock causes a depreciation of the US dollar as well as economic recession and excessive inflation in the global economy. The way in which monetary policy deals with the shock matters for the long-run level of economic activity. An unexpected US monetary tightening causes an appreciation of the US dollar and a fall in real GDP and inflation over the long run. The monetary policy reaction to this change seems to be stronger in developing countries than in developed countries. Simultaneous shocks to traded-sector productivity in China, India, Korea and Indonesia induce a rise in real GDP and currency appreciation in these four countries. Meanwhile, many Asian countries benefit from the shocks with higher productivity and GDP. The value of their currency is likely to appreciate.