The proposed studies will examine the relationship of mood and recall to attributional style. A variety of theories in personality, social and clinical psychology have suggested that cognitions play a role in determining mood. One set of cognitions that has received particular attention is that of causal attributions. For instance, attributional styl has been shown to interact with stress to predict subsequent depressive affect. However, little is know about the sources of attributional style. Pilot work has suggested that attributional style can be modified by having subjects recall different memories, though the effect of memories differs for relatively nondysphoric versus relatively dysphoric subjects. For nondysphoric subjects, recalling past failures led to even less depressogenic attributions that recalling past successes. For dysphoric subjects, on the other hand, recalling past failures led to even more depressogenic attributions, though this latter effect was nonsignificant with a relatively small sample. The four proposed studies will examine the boundary conditions for this interaction. The first will examine whether differences in the events recalled by dysphoric and nondysphoric subjects produced the differences in attributional style. In this study, the causes of the events that subjects process prior to making attributional judgement will be varied systematically. The second study will examine the implications of changes in attributional style for social comparison. This will be done by having subjects make attributional judgments both for self and other under different recall conditions. In the third study, subjects will recall events from either the interpersonal or academic arenas, but will make attributional judgments in both. This will permit examination of the generality of changes in attributional style. Finally, in the fourth study, the possibility that some people will display more general and other more specific changes in attributional style will be examined. This will b done by examining the differential effects of recall on the attributional style of subjects with more or less self-complexity.