Background: We set out to determine whether separable visual and representational components underlie normal subjects’ upward and distal biases in bisecting vertical and radial lines under visual guidance. Methods: Thirty-four normal subjects were asked to bisect lines oriented horizontally, vertically, and radially. Human silhouette figures were placed at either end of each line. These figures were presented upright or upside down in order to pictorially define a “top” to each line independent of the actual top of the visual field. Results: Although subjects erred toward the top of the visual field, they also demonstrated a significant bias toward the heads of the figures for lines in all spatial orientations. Conclusions: This result supports the existence of two biases: one toward the upper visual field, and another toward an internally represented “top” as suggested pictorially. These findings provide further support for the hypothesis that normal subjects’ upward and distal biases on bisection of vertical and radial lines under visual guidance have both representational and visual-based components.