Airline markets do not satisfy the theoretical conditions for perfect contestability, and theory does not indicate which deviations from perfect contestability significantly affect performance. Thus, the effect of potential entry on performance in airline markets is explored empirically. The best model for explaining fares is found to be one in which the measure of market concentration takes into account not only the number and size distribution of incumbents, but also the number of potential entrants not significantly disadvantaged due to economies of scale and scope. The hypothesis that all city-pair markets are perfectly contestable is rejected.