In recent years an emergent global policy discourse has promoted the concept of Integrated Water Resources Management (IWRM) as a strategy for the sustainable management of international waters. However, integration remains a considerable challenge in large international river basins. This article addresses the relevance of the global discourse on IWRM, asking how much integration can be expected and how much integration is desirable in international water management. The article presents a novel compilation of eighty-six international river basin organizations and examines their degree of integration in terms of three dimensions: membership, substantive scope, and form. More particularly, the article examines the integration problem from an economic perspective, asking whether integration serves the self-interest of the respective riparian states. The empirical evidence highlights the difficulties of integration, as the majority of international river basin organizations remain narrow in membership and scope. Economic considerations suggest that voluntary cooperation in river basins is institutionally demanding and that the degree of integration depends on the problem at hand. Hence, the challenge for international waters management is to search for the economically desirable degree of integration in each case. Keywords: international river basins, Integrated Water Resources Management, Coase theorem, integration. The Problem Some 263 river basins on earth cross international boundaries. (1) Whenever a water resource is international, competing claims over shared waters can generate conflict among riparian countries. In response, a global policy discourse has emerged that promotes Integrated Water Resources Management (IWRM) as a concept for sustainable resource use. IWRM calls for the integrated management of different water uses within the hydrological confines of a river basin. Proponents of IWRM often advocate the establishment of special river basin organizations. However, a number of social scientists have rejected IWRM as unrealistic with respect to international waters. For example, Frank Marty, basing his predictions on experiences with six bilateral agreements, expects that the majority of international water agreements will be bilateral and narrow in scope and that this is indeed how it should be. (2) John Waterbury warns that the quest for integrated development in international river basins will be elusive and excessively costly, although he concedes that bilateral agreements are not necessarily a solution for multiparty basins. (3) Empirical research suggests that existing arrangements are limited in membership and substantive scope, but large-N analyses are rare and generally limited to the analysis of membership. In terms of membership, Jesse Hamner and Aaron Wolf find that 86 percent of 145 international water treaties concluded since 1945 are bilateral; however, they do not distinguish between bilateral and multilateral basins. (4) According to Ken Conca, Fengshi Wu, and Ciqi Mei, thirty-three of forty-nine international river agreements established in multilateral basins between 1980 and 2000 are bilateral. (5) In terms of substantive scope, Ludwik Teclaff suggests that the majority of international water institutions are single-purpose, but the basis of his analysis is unclear. (6) In terms of form, it is an open question as to what extent riparian states not only conclude agreements but also set up specific river basin organizations. (7) Hence, only vague knowledge is available concerning integration of existing international river basin institutions in terms of substantive scope and form. How much integration can be expected in the management of international waters? How much integration is desirable in terms of membership, substantive scope, and form? In other words, how viable is the IWRM concept for the management of international rivers? To address these questions, I discuss in this article IWRM in the light of empirical evidence and economic considerations. …