Does reading fiction improve our ability to understand one another? Correlational data suggest that lifetime fiction exposure is positively associated with social outcomes. The latest experimental data suggest that fiction reading may slightly improve social ability, although this conclusion is tenuous. Here, we test fiction’s putative causal impact on social outcomes by conducting a randomized controlled study in which adult participants (N=210) were randomly assigned to engage in no reading for pleasure, or to read fiction or nonfiction for 45 min/day, five days/week, for four weeks. At the end of the study, participants were assessed on three classes of social outcomes: theory of mind (ToM), empathy, and social functioning. Contrary to other experimental work, fiction readers did not outperform nonfiction readers or participants who abstained from pleasure reading on any social outcome. Nonfiction readers outperformed those who abstained from pleasure reading on empathy measures. For fiction readers, narrative transportation was positively associated with empathy post-reading, and intrinsic motivation was positively associated with empathy and ToM post-reading. Contrary to other correlational work, we did not observe associations between lifetime fiction exposure and social outcomes. These data are consistent with the following possibilities regarding fiction’s positive impact on social outcomes: such findings may reflect a priming effect, may occur only after prolonged exposure to fiction, and/or may occur for readers who exhibit a particular kind of engagement with the reading.