The 1990s will be remembered in economic history as a decade of currency crises. In September 1992 the exchange-rate mechanism of the European Monetary System came under attack. This was followed by the Mexican currency crisis of December 1994 and, more recently, the Asian and Russian crises. 1 Concern about the possibility of these crises spreading to other countries and the implications this might have for the conduct of monetary policy rekindled interest amongst economists and policy-makers in the determinants of currency crises. There is no consensus in the theoretical literature regarding the causes of such crises. Traditional models suggest that currency crises are caused by deteriorating economic fundamentals, while more recent models link crises to self-fulfilling prophecies and contagion effects. Since different factors are identified as the causes, it is necessary to examine their determinants empirically.This article examines the determinants of currency crises in developing countries. It asks two basic questions: (a) are currency crises linked to economic fundamentals? and (b) is there any evidence of a contagion effect after controlling for the potential effects of economic fundamentals? Using a panel of annual data for 19 developing countries spanning the period 1977-97, it demonstrates that, among the macroeconomic fundamentals considered as predictors of currency crises, the current account deficit is the only variable that can be consistently linked to currency crises. Economic fundamentals such as the growth rate of domestic credit, lending booms, inflation, foreign debt, output growth and high fiscal deficits are generally not significant. In cases where a