An influential literature has emerged around the premise that there exists an uneasy tension between (1) bank-firm relationships that promote incentives for firm-specific investments by banks, and (2) competition between banks that can destroy such incentives. This paper studies the industrial organization of the investment banking market in order to shed light on how this tension may be resolved. This market is ideally suited to study this question because there is a vast literature establishing the fact that it is characterized by both relationships and competition. The model studies the impact on relationships of four di.erent faces of competition: nonexclusive relationships between banks and firms, competition from arm`s-length banks, non-price competition, and endogenous entry. The key premise of the model is that relationships involve sunk and non-verifiable costs. The first set of implications provide «possibility» results, which show how relationships are sustainable in the face of each of the four types of competition. Further, banks are shown to establish relationships without either local or aggregate monopoly power. A second set of results yields predictions on several characteristics of the observed market structure. Vertical segmentation, invariance of market concentration to market size in the relationship segment, and a competitive fringe that coexists with a stable oligopoly in equilibrium, characterize market structure. The model is applied to study the e.ects of global competition, and to provide a logic for antitrust analysis.