In this article we report meta-analyses of the relation of attributional styles to depression. In 104 studies involving nearly 15,000 subjects, several attributional patterns had reliable associations with depression scores. For negative events, attributions to internal, stable, and global causes had a reliable and significant association with depression. Studies in which the attribution factors of ability and luck were measured also showed a reliable association with depression. For positive events, attributions to external, unstable, and specific causes were associated with depression. Ability and luck attribution factors for positive events were also associated with depression. The relations for positive events, however, were weaker than the corresponding ones for negative events. In general, these patterns of relations were independent of a number of potential mediators suggested by authors in this literature, including the type of subject studied (psychiatric vs. college student), the type of event about which the attribution is made (real vs. simulated), the depression measure used, or the publication status of the research report. These conclusions are compared with those of other reviews. Implications for attributional models of depression are discussed.