Recently, we reported that group subjective measures of presence as well as observers' postural responses are sensitive to increasing the realism of a display with motion content, by the addition of stereoscopic information, using a 20-inch stereoscopic screen with an effective horizontal field of view of 28 deg. (Freeman, Avons, Meddis, Pearson, & IJsselsteijn, 2000). The experiment presented here employed a large projection display with a 50 deg. horizontal field of view showing a rally car traversing a curved track at speed. The independent variables included image motion and stereoscopic presentation as within-subjects factors and screen size as a between-subjects factor. Dependent variables included subjective measures of presence, vection, involvement, and sickness, as well as observers' lateral postural responses, which served as a candidate objective corroborative measure of presence. Results demonstrated a noisy yet positive effect of stereoscopic presentation on the lateral postural responses. Post-test subjective ratings revealed a significant effect of stereoscopic presentation on the subjective judgments of presence, but not on those of vection, involvement, or sickness. Image motion had a large and significant effect on the subjective judgments of presence, vection, and involvement. The effect of image motion was considerably larger than that of stereoscopic viewing. By comparing results between experiments, a large effect of screen size on subjective presence ratings could be demonstrated, but only for the video stimulus that contained motion. The postural response measure did not differentiate between screen sizes, thus limiting its utility as an objective corroborative measure of presence, although further research is required to be able to be more firm in our conclusion regarding this issue.