As in international tests of purchasing power parity, new more powerful univariate unit root tests have been successful in rejecting a unit root process in US city relative prices over the period 1918-1997. However, convergence rates calculated with these tests are at odds with theories of price convergence. This article addresses structural breaks that may be misinterpreted as nonstationary permanent stochastic process. Using unit root tests with structural breaks, we are able to reject a unit root process in the majority of US city relative prices over the entire sample period. Moreover, convergence rates fall to those more in line with theory.