We analyze a fundamental, but so far unaddressed, aspect of leadership and social comparison: how leadership can influence the employee social comparison processes inherent to performance contexts such that they are more conducive to subsequent performance. Based on the observation that people tend to compare in counterproductive ways, which detracts from future performance, we advance and test theory on how leaders can improve employee performance by influencing the social comparison processes between them. Results from a randomized field experiment in a contact center with objective performance data from 290 employees show that leadership focused on influencing assimilative social comparison improved performance in the five months after the intervention. As hypothesized, this effect was mediated by employee assimilative social comparison and was only obtained when leaders created a collaborative context in which to influence the assimilative social comparison. Our theory and findings have important implications for both leadership and social comparison research.