Executive function mechanisms underpinning language-related effects on theory of mind understanding were examined in a sample of 165 preschoolers. Verbal labels were manipulated to identify relevant perspectives on an explicit false belief task. In Experiment 1 with 4-year-olds (= 74), false belief reasoning was superior in the fully and protagonist-perspective labeled conditions compared to the child-perspective and nondescript labeling conditions. In Experiment 2 with 3-year-olds (= 53), labeling the nondominant belief only biased attentional inertia. In Experiment 3 testing generalization in 4-year-olds (= 38), labeling manipulations translated to improved performance on a second label-free explicit false belief task. These outcomes fit a cognitive flexibility account whereby age changes in the effects of labeling turn on formulating sophisticated conceptual representations.