Previous research has shown that search and rescue robots get stuck. We ask why. Perhaps operators attempt to drive through apertures larger than the robot, but too small to be driven through unhindered. If this is the case, then operators should base their decisions to enter apertures on their ability to drive the robot instead of other physical dimensions. This assumes that operators are cognizant of their abilities to drive the robot through the aperture. To test this assumption, participants viewed an image transmitted from a camera mounted on a robot and drove towards apertures of varying width. Half of the participants judged whether the robot could pass through, half judged whether they could drive the robot through. Finally, all participants attempted to drive the robot through the aperture. Results indicated that pass-ability judgments were accurate, but drive-ability judgments were not. Implications for training and interface design are discussed.