This paper investigates the long-run purchasing power parity hypothesis when exchange rate returns and inflation rates are assumed to be heavy-tailed stochastic processes. More specifically, residual-based and likelihood-ratio-based cointegration tests of PPP that explicitly allow for infinite-variance innovations are applied to monthly data (1973:1-1999:12) for Belgium, Canada, Denmark, France, Germany, Italy, Japan, the Netherlands, Norway, Spain, Sweden, and the United Kingdom. Our test results are marginally less supportive of PPP when the innovations are assumed to be infinite-variance, α-stable processes.