We provide such a conceptualization and report evidence bearing on its implications. In two experiments, college students either imagined themselves participating in an audiotaped get-acquainted situation in which the relative amount of time that they and their partner spent talking was varied. The relative attention they paid to the statements they personally made and the statements their partner made was ma-nipulated either by instructions or by varying the partner's physical at-tractiveness (thus motivating them either to form an impression of their partner or to create a good impression of themselves). In a third exper-iment, participants actually engaged in a get-acquainted conversation with one another. All three experiments confirmed the assumption that people's immediate and retrospective estimates of a conversation's duration, and their reactions to the conversation, are influenced in opposite ways by their objectives at the time the conversation took place and the aspects of the interaction to which they attended.