Hypothesized that rating bogus "personality feedback" and answering personality inventory items are both instances of the same general behavior. After undergraduates were administered personality inventories, they were asked to evaluate the accuracies of personality descriptors under differing instructional sets. In support of the hypothesis, a Barnum Group's (N equal to 24) personality inventory responses and "personality feedback" ratings correlated significantly and as highly as Reliability Controls' (N equal to 24) alternate forms reliability coefficient. Inventory responses and "feedback" ratings were affected equally by the descriptors' favorability and by the Ss' defensiveness. Contrary to the hypothesis, descriptors were rated as more personally accurate when presented as "feedback" than when presented as test items. Theoretical and applied implications of the findings are discussed.