In The lord of the rings: the fellowship of the ring, Boromir refers to the lands of Mordor as the place where evil never sleeps. Cinematic evil itself never sleeps, always arising in new forms, to the extent that there exist as many types of evil as there are films. This thesis examines this constantly shifting construction of evil in American popular cinema between 1989 and 2002--roughly, the period between the fall of the Berlin Wall and the attack on the World Trade Center--and how this cinema engaged with representations of enemies and of evil per se. The thesis uses content and thematic analysis on a sample of the 201 most successful films at the U.S. box office during the period. In these films, cinematic evil is constructed according to a visual aesthetic that attempts to engage with societal values, but fails to do so due to the emphasis on its visual construction and its commodification. As Baudrillard argues, evil has become a hollow concept devoid of meaning, and this is especially so for cinematic evil. It is recognised, and is recognisable, by the visual excessiveness of its violence (or potentiality for violence), and by certain codes that are created in reference to intertextual patterns and in relationship to discourses of paranoia and malaise. But cinema in this period failed to engage with the concept of evil itself in any meaningful way. Cinematic evil mirrors the descent into the chaos and disorder of a postmodern society. All cinematic evil can do is to connect with this sense of unease in which the "reality of evil" cannot be represented. Instead, it draws on earlier icons and narratives of evil in a conflation of narrative and spectacle that produces a cinema of nostalgia. Moreover, stripped of narrative causality, these films express a belief, unproved and unprovable, that evil things and evil people may arise in any form, in any place and at any time: a cinema of paranoia. Together, these factors produce a cinema of malaise, perpetually confronting an evil it is unable to define or locate.