This study presents the significance of the currency crises, discusses the related literature and applies a model of economic vulnerability to Turkey during 1985Q2-2004Q2. The common approach in currency crisis literature is to focus on the performance of thresholds for a set of early warning indicators. Following the explanation of “Index of Speculative Pressure” (ISP), Granger causes of the ISP is discussed. The study shows that, current account/ GDP ratio, M2/international reserves ratio, real credit growth and current account/foreign direct investment ratio are Granger causes of the ISP at 1% level. Then by using Vector Auto Regression (VAR) model, the ISP index is forecasted. The study shows that the combination of VAR(1)+VAR(2)+VAR(5) models generate relatively better forecast values than all other single models. Finally the study estimates dynamic probit and logit models by using maximum likelihood to predict currency crises. It shows that logit model gives a better performance than the probit, for a better prediction of the probabilities of the Turkish currency crises. The most important contribution of this study is to show that the logit model has a very high performance in the prediction of Turkish currency crises. It can be used to foresee forthcoming currency crises. Also the forecast of the ISP (as a level) is giving very successful results. It is observed that the ISP and forecasted ISP values are almost moving together or very close to each other.