This research stems from the casual observation that the image of a television screen with 18° geometric ®eld of view (FOVg) as seen on a real television with 17° real FOV appears much “nearer” than the real television<Eth>as much as 98% closer. Does the image appear “nearer” because the distance to the real television is misjudged or is a mental model of a virtual subjective self-location created? Either way, all the projective mapping in the world, whether only in the z plane or in all x, y, and z planes, is irrelevant to explain this powerful perceptual or cognitive effect that the neglected variable of FOV has on self-location in virtual space. Accurate perception of the scene and precise self-location in virtual environments is the goal of accurate perspective geometries, scene computation, and helmet-mounted display (HMD) optics. Yet, as research has already clearly shown, accurate geometric projection is no guarantee of accurate perception: for instance, images viewed exactly at their proper projection points have repeatedly been seen at distorted distances or inaccurate directions. In addition to precise engineering, it is important to understand the psychology of selflocation, also called egocenters, to obtain an understanding of virtual space. A new theory, Cognitive Frame Theory, is proposed to deal with cognitive modi®cations of perception in a way that emphasizes the importance of self-location. Cognitive Frame Theory builds on the phenomenal geometry underlying self-location perceptions: the localization of objects in space requires a combination of perceived distance, perceived direction, and the perception of the position or motion of the self. A fundamentally novel component of the theory suggests that observers use their natural ®elds of view of 120° vertical by 180° horizontal as the primary standard for interpreting visual displays and their self-location in the space of those displays.