Being observed magnifies actions and their perceived impact: I test the hypothesis that people, when observed, perceive their actions as more substantial because they add the audience’s perspective to their own perspective. Participants who were observed while eating (Study 1) or learned they were observed after eating (Study 2) recalled eating a larger portion than unobserved participants. The presence of others magnified both desirable and undesirable actions. Thus, observed (vs. unobserved) participants believed they gave both more correct and incorrect answers in a lab task (Study 3) and, moving to a field study, the larger the audience, the larger the contribution badminton players claimed toward their teams’ successes as well as failures (Study 4). In contrast to actions, inactions are not magnified, because they are unobservable; indeed, observed (vs. unobserved) participants believed they solved more task problems but did not skip more problems (Study 5). So far, I have measured perceptions of behavior. Building on these findings, I next investigate whether people also perceive the impact of their actions to be smaller versus larger when observed. I find that when observed, people think the impact of easy actions is larger, but the impact of difficult actions is smaller. Taken together, these studies show that being observed fundamentally alters the subjective magnitude of one’s actions, and characteristics of the observer can alter the behavior itself.