A small experimental, central scotoma significantly attenuates the human pattern visual evoked potential. The steady-state pattern visual evoked potential was recorded from seven visually normal adults who viewed a reversing checkerboard with 24' checks and a central scotoma that varied in size and shape. We found that square scotomas had to be at least 3 x 3 degrees to significantly (p < 0.05) attenuate the pattern visual evoked potential. Receptor density has been shown to be greater along the horizontal meridian than the vertical meridian. We hypothesized that this results in greater cortical representation of the horizontal meridian than the vertical meridian and, therefore, the pattern visual evoked potential might be significantly attenuated by a smaller rectangular scotoma oriented along the horizontal meridian than along the vertical meridian. One dimension of the rectangular scotoma was fixed at either 1 degree or 3 degrees, while the other dimension was varied from 1 degree to 8 degrees. The threshold scotoma size that significantly (p < 0.05) attenuated the pattern visual evoked potential was a horizontal scotoma subtending 1 x 4 degrees and a vertical scotoma subtending 5 x 1 degree (vertical x horizontal). Meridional differences in cortical representation were not apparent to the larger scotoma series in which the fixed dimension subtended 3 degrees (3 x 2 degrees and 2 x 3 degrees). Further analysis of the data revealed that the apparent meridional difference for the 1 degree scotoma series was a function of data variability. The determinant of the PVEP amplitude was scotoma area, not orientation. Monocular and binocular threshold scotoma sizes were the same, which could be due to the level of binocular summation demonstrated by our subjects.