["Journal of Applied Social Psychology, EarlyView. ", "\nABSTRACT\nPerspective‐taking is a popular method for reducing prejudice in research and practice, but it can elicit reactance and exacerbate hostility. Why do we observe such varied responses when people are invited to engage in perspective‐taking for marginalised outgroups? We suggest that people do not enter experimental settings as blank slates, and the literature has little considered how existing (a priori) perceptions of specific social groups may impact motivation to engage in perspective‐taking. Two cross‐sectional studies examine how perceptions of the target group and threat relate to willingness and ability to engage in perspective‐taking. In Study 1 (N = 213), we found that group perceptions (i.e., warmth, competence, and morality), and threat (dissimilarity, realistic, and symbolic threats) were the primary predictors of willingness and ability to engage in perspective‐taking. Study 2 (N = 886) used multigroup structural equation modelling to demonstrate that perceptions of warmth, competence, and morality were associated with lower levels of dissimilarity, symbolic, and realistic threat. In turn, heightened dissimilarity threat was negatively associated with willingness and ability to perspective‐take, whereas increased symbolic threat was positively associated with ability to perspective‐take. Findings suggest that the way people perceive groups, and the related feelings of threat, are associated with (dis)engagement in perspective‐taking. The results highlight key considerations for the application of this popular prejudice‐reduction strategy.\n"]