In this paper we aim to analyze the long-run validity of the Purchasing Power Parity (PPP) hypothesis for the Romanian exchange rate. Our goal is achieved using Zivot-Andrews test with one structural break in order to identify changes in real exchange rate compared with traditional tests like Augmented Dickey-Fuller and Phillips-Perron and cointegration analysis in order to identify the long-run relationship between exchange rate and domestic and foreign prices. Real exchange rate stationarity implies that a shock it is absorbed in time and PPP holds in long-run. If nominal exchange rate and prince indices are non-stationary we verify if the variables are cointegrated as PPP weak form and symmetry and proportionality conditions as PPP strong form. We identify evidence of cointegration for all three models, but we don’t find any evidence to support symmetry and proportionality condition for PPP strong form case. Also, we use three different price indices:consumer price index, consumer price index without regulated prices and industrial producer price index in order to identify which indices is more relevant for our analysis. The monthly data cover the 2001M01-2011M09 period. The empirical analysis provided mixed results depending on the used price index and methodology.