Some empirical and theoretical studies have emphasized on fixed exchange rate regime in controlling time inconsistency, while others consider the role of target zone regime as an important factor. Thus there is no general consensus to decide which exchange rate regime may bring about less time-inconsistency. The main purpose of this study is to investigate policy time-inconsistency in exchange rate regimes of Iran (controlled exchange regime between 1991-2001 and managed Floating between 2002-2012). For this purpose, a fuzzy prediction system is used which includes three steps for estimating time-inconsistency. In the first step, fuzzy C-Means is used to calculate membership functions. Then Takagi Sugeno Model (TSM) is applied to identify the model structure, and finally Recursive Least Squares (RLS) is conducted to estimate the amount of time-inconsistency. The results show that the amount of policy time inconsistency under managed Floating exchange rate regime is less than controlled exchange rate regime. Thus it is suggested to policy makers to continue managed Floating exchange rate regime, so that the fluctuations of output decreases. At the same time, since managed Floating exchange rate regime imposes more price fluctuation on economy it is necessary for the policy makers to pay enough attention to the issue.