This paper investigates the long- and short-run determinants of the real exchange rate using a panel of data for fourteen OECD countries. The data are analyzed using time series and panel unit root and panel cointegration methods. Two dynamic productivity-based models are used to motivate the empirical exercise. The candidate determinants include productivity levels in the traded and in the nontraded sectors, government spending, the terms of trade, income per capita, and the real price of oil. The empirical results indicate that it is easier to detect cointegration in panel data than in the available time series; moreover, the estimate of the rate of reversion to a cointegrating vector defined by real exchange rates and sectoral productivity differentials is estimated with greater precision as long as homogeneity of parameters is imposed upon the panel. It is more difficult to find evidence for cointegration when allowing for heterogeneity across currencies. The most empirically successful model of the real exchange rate includes sectoral productivity measures in the long run relation and government spending in the short run dynamics.