We examined the utility of two measures of cognitive bias—a self-report questionnaire, the Dysfunctional Attitudes Scale (Weissman and Beck
1978), and a cognitive processing task, the Scrambled Sentences Test (Wenzlaff
1988,
1993), in predicting episodes of depression prospectively in a community-based sample of nondepressed women. When examined separately, each measure contributed significantly to predicting a subsequent diagnosis of MDD, supporting the utility of both types of measures as indicators of depression vulnerability. When examined together, each measure made an independent contribution to the prediction of MDD, raising the possibility that questionnaire and processing tasks may assess separate aspects of cognitive vulnerability to depression.