Discussions surrounding the emotional impact of film have largely been the domain of film theory and criticism. And while psychologists have ventured into the empirical study of cinematically-induced emotion, such work has traditionally focused on the narrative elements of film, most often in the context of isolated film clips. The current work examines the influence of low-level cinematic structure on emotional response among viewers in the context of the whole film. Specifically, this dissertation investigates the emotional impact of clutter, luminance, motion, shot density, shot scale, and sound amplitude across four short films. These low-level metrics are objectively quantifiable and provide a backdrop against which to analyze the time course of emotional response as a film unfolds. Emotional response is captured via three distinct measures of emotional tension—joystick displacement as a subjective measure, and heart rate and electrodermal activity as objective measures. An analysis of 40 viewers reveals, first and foremost, that emotional tension is synchronized within all three emotion measures, demonstrating that people respond in remarkably similar ways to cinematic narratives. The low-level structure of film at least partially explains this synchrony, as all six cinematic variables predicted emotional tension in at least half of the short films utilized in this study. An omnibus analysis reveals that film structure exerted the most robust effect on subjective emotional tension, with all variables but clutter explaining a substantial portion of variance in joystick displacement. Subjective emotional tension was positively predicted by motion, shot density, shot scale, and sound amplitude, and negatively predicted by luminance. Given that these findings are limited to films featuring White characters, and that group-based perceptual biases might disrupt emotional engagement with more diverse characters on screen, a second study was carried out to enable expansion of this work on a corpus of racially diverse facial stimuli. A collection of 2048 images was digitally rendered and validated for racial representativeness across four geographical regions. Future directions and social implications of this work are discussed, with a focus on applications to intercultural communication and museum science, as well as the translation of the current findings to broader contexts.