The financial crises of the 1990s have created the perception that one of the fundamental reasons for the occurrence of such crises is to be found in the fact that exchange rates were pegged for too long. These pegged exchange rates inevitably invited speculative attacks in the foreign exchange markets that quite often spilled over to the banking sector, and led to banking crises. Recently, the analysis of the financial crises has led to a new consensus among policy makers, i.e., the bipolar view (see Fischer, 2001). According to this view, countries should allow for either flexible or irrevocably fixed exchange rates in order to avoid future crises while the intermediate solutions, such as pegged exchange rate regimes, should be avoided. The advantages of such intermediate exchange rate regimes have been offset by the disadvantages in terms of uncertainty in the financial markets. In this paper we analyse this link between the exchange rate regime and the probability of financial crises. We first analyse in the next section the relation between the exchange rate regime and the occurrence of foreign exchange crises, while Section 3 briefly reviews the associated empirical evidence. We then study the relation between the exchange rate regime and the occurrence of banking crises (Section 4). This analysis will also allow us to connect crises in the foreign exchange markets and banking crises. Section 5 concludes.