We use several popular tests to test the validity of the Purchasing Power Parity (PPP) hypothesis. In particular, we analyze four classes of tests-standard univariate unit root tests, co-integration, panel unit root tests, and unit root tests for nonlinear frameworks-for a dataset consisting of 20 bilateral exchange rates. Through this approach, we ascertain the effectiveness of each methodology in assessing the validity of PPP. Overall, our results suggest little evidence to support PPP. Among the conducted tests, the Panel Analysis of Nonstationarity in the Idiosyncratic and Common components (PANIC) provides the richest insights by disentangling the possible sources of non-stationarity of real exchange rates. The relevance of using price indices with different characteristics is also pinpointed.