This research examines the developmental origins of a form of partiality that can threaten the integrity of merit-based evaluations. In each of two studies, 3- to 6-year-old children (total N = 335) rated drawings made by child artists they had never met, with the degree of identification between evaluator and artist manipulated across trials. In Study 1, ratings were systematically inflated when the evaluator and artist were identified to each other, but not when only the artist was identified to the evaluator. Study 2 showed that merely identifying the evaluator was sufficient to produce inflated ratings. These results show that anonymity has implications for making fair evaluations as soon as early childhood, and that self-relevant considerations can affect how children evaluate the work of others.