The assumption that the Ebbinghaus illusion deceives perception but not grasping, which would confirm the two visual system hypothesis (TVSH) as proposed by Milner and Goodale (1995), has recently been challenged. Franz and colleagues (2003) found that the illusion affects both perception and grasping, and showed that the motor effect cannot be accounted for by a selective obstacle avoidance mechanism, which assumingly affects grip scaling only for wide gaps between the central target and context circles. We therefore compared kinematic parameters of prehension, using the Ebbinghaus illusion, and a neutral control condition where subjects (n=11) grasped a disc without any context circles. In accordance with the perceived size, the peak grip aperture (PGA) was smaller when the disc was surrounded by large context circles, and larger for small context circles. However, the largest PGA values were found in the control condition. Hence the presence of context circles generally leads to a reduction of the PGA. This reduction effect casts doubts on how appropriate it is to compare perceptual and motor effects using the Ebbinghaus illusion. Since the PGA is reduced by context circles, even smaller motor effects in comparison to perceptual effects cannot be unequivocally interpreted in favour of the TVSH. Hence, since such motor effects cannot be separated from the reduction effect, we have to conclude that this illusion should not be used when testing the TVSH.