Research on self-esteem has focused almost exclusively on level of trait self-esteem to the neglect of other potentially more important aspects such as the contingencies on which self-esteem is based. Over a century ago, W. James (1890) argued that self-esteem rises and falls around its typical level in response to successes and failures in domains on which one has staked self-worth. We present a model of global self-esteem that builds on James' insights and emphasizes contingencies of self-worth. This model can help to (a) point the way to understanding how self-esteem is implicated in affect, cognition, and self-regulation of behavior; (b) suggest how and when self-esteem is implicated in social problems; (c) resolve debates about the nature and functioning of self-esteem; (d) resolve paradoxes in related literatures, such as why people who are stigmatized do not necessarily have low self-esteem and why self-esteem does not decline with age; and (e) suggest how self-esteem is causally related to depression. In addition, this perspective raises questions about how contingencies of self-worth are acquired and how they change, whether they are primarily a resource or a vulnerability, and whether some people have noncontingent self-esteem.