Many scholars and practitioners claim that labeling groups or individuals as “terrorists” does not simply describe them but also shapes public attitudes, due to the label's important normative and political charge. Yet is there such a “terrorist label effect”? In view of surprisingly scant evidence, the present article evaluates whether or not the terrorist label—as well as the “Islamist” one—really impacts both the audience's perception of the security environment and its security policy preferences, and if yes, how and why. To do so, the article implements a randomized-controlled vignette experiment where participants (N = 481) first read one out of three press articles, each depicting a street shooting in the exact same way but labeling the author of the violence with a different category (“terrorist”/“shooter”/“Islamist”). Participants were then asked to report on both their perceptions and their policy preferences. This design reveals very strong effects of both the “terrorist” and “Islamist” categories on each dimension. These effects are analyzed through the lenses of social and cognitive psychology, in a way that interrogates the use of the terrorist category in society, the conflation of Islamism with terrorism, and the press and policymakers' lexical choices when reporting on political violence.