---

Inflation is a major problem in many developing countries and high levels of inflation leads to economic growth reduction, misallocation of resources and relative prices disorder. Since linear models have been used in most of the for time series modeling, so this paper using a nonlinear model aims to increase the empirical literature of the inflation sustainability in Iran. This study using the nonlinear STAR models examines the hypothesis that the monthly inflation rate has been stable during 1962-2007. In order to study the unit root hypothesis the Ng, Perron and KPSS testes also have been used. Results suggest that the monthly inflation rate in Iran is nonlinear and follows the LSTAR model. Results also indicate that the monthly inflation rate is locally unstable, so in low regime (less than 2/5 percent) is stable and in a high one is unstable and whenever inflation is in high regime the cost of each inflation rate control policy to is more than it’s benefit, but stability analysis based on unit root tests indicates that monthly inflation rate in the economy of Iran has long-term structure