Two experiments examined the effects of pictorial realism, observer interactivity, and delay of visual feedback on the sense of “presence.” Subjects were presented pairs of virtual environments (a simulated driving task) that differed In one or more ways from each other. After subjects had completed the second member of each pair they reported which of the two had produced the greater amount of presence and indicated the size of this difference by means of a 1-100 scale. As predicted, realism and interactivity increased presence while delay of visual feedback diminished it. According to subjects' verbal responses to a postexperiment Interview, pictorial realism was the least influential of the three variables examined. Further, although some subjects reported an increase in the sense of presence over the course of the experiment, most said that it had remained unchanged or become weaker.