International institutions are a central focus of international relations scholarship as well as of policymaking efforts around the world. Despite their importance, our scholarly literature lacks a widely accepted definition of just what they are. Instead, scholars have employed a range of largely nonoverlapping conceptions, contributing to a fragmentation of the literature and hindering theoretical cumulation. This essay seeks to remedy this unsatisfactory state of affairs. It first reviews the principal ways in which international institutions have been conceptualized and identifies their shortcomings. It then develops a definition that promises to be inclusive of what are commonly regarded as the most important institutional forms without losing analytical coherence. A final section discusses some of the concrete benefits that result from employing the new definition, both in improving existing scholarship and by suggesting valuable new avenues of research.