Content features of persuasive messages modulate neural activity and this variation can be used to identify messages that are most likely to result in attitudinal and behavioral change. Informed by the Elaboration Likelihood Model of persuasion, my research applies a “brain-as-predictor” approach to understand how audiences evaluate counter-attitudinal messages. In this study, I exposed subjects that were either at high- or low-risk for drug use to anti-drug public service announcements (PSAs) while undergoing functional magnetic resonance imaging. My results show that neural activity in brain regions associated with executive and self-referential processing accurately predict perceptions of message effectiveness in independent samples. Moreover, this prediction accuracy exceeds that achieved by traditional self-report measures. Previous efforts have struggled to identify messages that resonate with high-risk audiences as these viewers tend to rate anti-drug PSAs as having low effectiveness regardless of variation in the message. My work demonstrates that it is possible to use neuroimaging data to overcome this obstacle.