When an individual grasps a physical object, the visuomotor system is able to specify grip aperture via absolute visual information. In contrast, grasping to a location previously occupied by an object (i.e., pantomime-grasping) results in the specification of grip aperture via relative visual information. The basis for the dissociable visual codes is thought to reflect that pantomime-grasping is a perception-based task. It is, however, important to recognize that grasping a physical object and pantomime-grasping differ not only in terms of their visuospatial properties but also with respect to the availability of haptic feedback: Grasping a physical object provides haptic cues derived from touching the object, whereas no such feedback is available in a traditional pantomime-grasping task. As such, we examined whether haptic feedback influences the information supporting a pantomime-grasp performed after a 1000-ms visual delay. Participants completed responses in each of the three tasks: (1) grasping a physical object, (2) traditional pantomime-grasping wherein the to-be-grasped object was removed from the grasping environment and (3) pantomime-grasping wherein the experimenter placed the object between participants' thumb and forefinger once they had completed their response (i.e., pantomime-grasping with haptic feedback). Just-noticeable-difference (JND) scores were computed to determine whether responses adhered to or violated the psychophysical (i.e., relative) principles of Weber's law. JNDs for the traditional pantomime-grasping task adhered to Weber's law, whereas JNDs for grasping a physical object and for pantomime-grasping with haptic feedback violated the law. Thus, we propose that convergent visual and haptic cues support the absolute specification of object size in a pantomime-grasping task. Furthermore, our results highlight the important role of multisensory cue integration in a target-defined grasping task.