As the popularity of virtual reality as an exposure therapy increases, it is important to validate the use of computer-generated stimuli in comparison to standardized images of “real” phobic objects, such as those of the International Affective Picture System (IAPS). The present study examined physiological and subjective measures of negative affect when viewing static IAPS images, static computer-generated images and moving videos of computer-generated images of feared stimuli and other negative stimuli which were not specifically feared. For example, a picture of a spider would be a “feared” stimulus for a spider fearful participant, whereas a picture of a snake would be categorized as a “negative” stimulus for that participant. Eighteen participants scoring high (high fear (HF) cohort) on questionnaires assessing specific fears of spiders or snakes and 20 participants scoring low (low fear (LF) cohort) on the questionnaires viewed the stimuli. The computer-generated videos elicited greater physiological (skin conductance and startle eyeblink potentiation) and self-report arousal responses than the IAPS images and the computer-generated static images. Computer-generated stills and IAPS images did not differ in eliciting emotional responses. Additionally, HF participants showed greater heart rate acceleration and larger skin conductance responses to their feared stimulus than to the negative stimulus, especially when viewing computer-generated moving videos. The results demonstrate the importance of motion in eliciting fear and the usefulness of computer-generated stimuli in the study of emotion.