Despite the increasing number of film rankings that are created on the basis of economic considerations, critical acclaim, or personal taste, cinema lacks the means to quantify the long-term success of films. The authors therefore propose a novel approach to the analysis of cinematic influences which is based on a film citation network that is extracted from a large data base. The resulting top list of films is more diverse in terms of the main creators, genre, actors, and technical specifications than a representative selection of personal favorite lists, voting lists, lists of individual experts, or lists deduced from expert polls.