An individual's peers at work affect her on-the-job performance in important ways. Prevailing research argues that the magnitude of these peer effects depends on the worker's or peers' absolute levels of performance. In contrast, theorize about the distance between the worker's performance and those of one's peers, rather than the absolute performance levels of either one. I argue that the extent to which a worker's performance changes in response to peers depends on how much better or worse one's peers are relative to the focal worker, which I term