This paper analyzes the stability of alternative exchange rate regimes in the face of substantial capital mobility. This issue often goes under the label of the unstable middle or the two-corners hypotheses. We argue that both the issues of why the middle is unstable and how far toward the extremes of fixed or flexible exchange rates countries need to go in order to substantially reduce the likelihood of currency crises depends crucially on political economy as well as technical economic considerations. We undertake a large N empirical study that extends the currency crisis literature by using a new classification of exchange rate regimes from the IMF and taking into account the interactive effects between government strength and alternative exchange rate regimes on the probability of currency crises. We find that weak governments increase the likelihood of currency crises under any type of exchange rate regime, and as our theory suggests, this effect is strongest under adjustably pegged exchange regimes.