Directional bias on cancellation has thus far not been standardized. While cancellation tasks are primarily used to assess lateral performance asymmetries, they may also reveal two-dimensional (i.e., combined lateral and radial) neglect patterns. We propose a method to evaluate and report cancellation neglect regardless of whether the neglect pattern is strictly unilateral or two-dimensional. Our method establishes the location of the geographic center of all neglected stimuli relative to the page center by averaging their Cartesian coordinates. This 'neglect center' is reported in polar coordinates to indicate its distance and direction from the page center. We apply our method to published examples of two-dimensional neglect. We find that neglect centers from different cancellation performances may not be statistically distinct even though they may occupy different quadrants. In addition, the net direction of neglect found by the coordinate method may differ from that inferred from measuring differences in quadrant omission totals. The suitability of the coordinate vs. the quadrant method will depend on the mechanism hypothesized for visuospatial exploration under particular test conditions. Using both approaches may detect different attentional biases operating during the same task. The coordinate method is appropriate for conventional cancellation testing. By incorporating the precise locations of all neglected stimuli and determining the net neglect direction in two dimensions, the technique may stimulate more comprehensive explanations for directional bias.