||Past research has obtained seemingly contrary findings regarding whether optimists, people who hold positive expectations about their future, approach or avoid self-relevant negative information. The present study (N= 178 college students) aimed to clarify the role of dispositional optimism by examining personal relevance of the information as a factor that potentially moderates optimists' and pessimists' preferences for information. Personal relevance of the information was operationalized in two ways - in terms of task importance, manipulated via information that a challenging verbal task measured an ability that was either highly important or relatively unimportant with respect to success in college, and task diagnosticity, manipulated via information that the test was either reliable in detecting the ability accurately or not. After completing the word task, all participants rated their interest in receiving feedback about their strengths and weaknesses on the task, and indicated their choice of feedback in case that they could receive only one kind of feedback. Due to their hypothesized ability to withstand the emotional costs of negative feedback when the feedback had the potential to be useful, it was predicted that optimists would select weakness feedback when the task was said to be both highly important and highly diagnostic, but would instead select strength feedback when the task was relatively unimportant and nondiagnostic. In the condition in which the feedback was most relevant (high task importance, high test diagnosticity), participants high in optimism were 44% likely to select weakness-focused feedback on the forced-choice measure, compared to 27% among participants low in optimism. In contrast, when the task was said to be highly diagnostic, but unimportant, only 14% of optimists chose weakness feedback. This pattern was not seen in the absolute ratings of interest in both strengths and weakness feedback, as most participants reported a high degree of interests in both types of feedback. Contrary to prediction, optimists did not differentially prefer strengths feedback when information was least relevant (low task importance, low test diagnosticity). Neither neuroticism nor beliefs in the malleability of one's performance on the task accounted for the relation between optimism and feedback preference. Limitations of the study and directions for future research are discussed.