This paper describes the use of a large screen virtual environment to induce the perception of translational and rotational self-motion. We explore two aspects of this problem. Our first study investigates how the level of visual immersion (seeing a reference frame) affects subjective measures of vection. For visual patterns consistent with translation, self-reported subjective measures of self-motion were increased when the floor and ceiling were visible outside of the projection area. When the visual patterns indicated rotation, the strength of the subjective experience of circular vection was unaffected by whether or not the floor and ceiling were visible. We also found that circular vection induced by the large screen display was reported subjectively more compelling than translational vection. The second study we present describes a novel way in which to measure the effects of displays intended to produce a sense of vection. It is known that people unintentionally drift forward if asked to run in place while blindfolded and that adaptations involving perceived linear self-motion can change the rate of drift. We showed for the first time that there is a lateral drift following perceived rotational self-motion and we added to the empirical data associated with the drift effect for translational self-motion by exploring the condition in which the only self-motion cues are visual.