In this article we test the purchasing power parity (PPP) hypothesis during the recent floating exchange rate period, using quarterly data for 21 OECD countries. In doing so, we use the long-horizon regression approach developed by Fisher and Seater [American Economic Review 83 (1993) 402] and consider 60 bilateral intercountry relations. We investigate the power of the long-horizon regression tests, using the inverse power function of Andrews [Econometrica 57 (1989) 1059], and provide weak evidence in favor of PPP.