The important concept of purchasing power parity (PPP) has a number of practical implications. Our central objective is to examine the stationarity of Turkeyâ€™s real exchange rates to test for the empirical validity of PPP. Our results from conventional univariate unit root tests fail to support PPP. However, when we use the empirical methodology developed by Caner and Hansen (2001), which allows us to jointly consider non-stationarity and non-linearity, we find evidence of non-linear mean reversion in Turkeyâ€™s real exchange rates. This implies that PPP holds in one threshold regime but not in another.