This paper attempts to determine the extent to which common external shocks explain simultaneous currency crises. We define crises on a country by country basis using a new criterion that takes into account variations in the volatility of exchange rates over time and across countries. Using a Poisson regression model, we find that over the post-Bretton woods period, a small number of common external shocks can explain between sixty to eighty percent of the variation in the total number of crises over time, depending upon the set of countries one looks at. Our findings provide one explanation of why currency crises sometimes bunch together and sometimes do not. * The authors thank Casey Cornwell and Mark Peralta for research assistance. Any opinions expressed in this paper are those of the authors and not necessarily those of the Federal Reserve Bank of San Francisco or the Federal Reserve System.