A modified version of the Oppel-Kundt illusion (i.e., a filled space is perceived as more expanded than an empty space of the same length) has been previously employed to distort space representation both in patients with neglect (i.e., failure to report/react to stimuli located in the space contralateral to the brain lesion) and in healthy participants. In those experiments, participants had to bisect or extend horizontal segments on backgrounds of exponentially spaced vertical lines. The exclusive use of visuo-motor tasks, however, did not allow excluding that the results were accounted for by a bias occurring at a response level of stimulus processing rather than by a visual illusion. To address this issue, in addition to a traditional line bisection task, a line length estimation task was employed, which allowed separating response and illusion-related factors. The results demonstrated that performance depended on the visual illusion rather than on a response bias. It was concluded that this version of the Oppel-Kundt illusion can be successfully employed to modulate space representation in humans.