The abstract below was accidentally omitted from the Wednesday posters session. It should have appeared in second position under the heading “Motion perception and temporal issues” (page 110). ♦ Motor and perception-based prediction A Sciutti, F Nori, G Metta¶, T Pozzo§, G Sandini (Istituto Italiano di Tecnologia, Genoa, Italy; § INSERM U887, Motricité Plasticité, France; e-mail: alessandra.sciutti@iit.it ) In this study we investigated whether prediction is differently realized when it is aimed at ‘driving a motor act’ and when instead its purpose is ‘perceptual-only’. In particular we focused on how dynamical information of target motion is used depending on prediction goal. In a first experiment a ball disappeared behind an occlusion after following parabolic paths. To analyze the ‘driving a motor act’ condition we asked subjects to move their arm to intercept the ball. In a second experiment we investigated the ‘perceptual-only’ condition by repeating a similar task but requiring only a perceptual evaluation. Subjects had to select (in 2AFC protocol) where the ball would have arrived after the occlusion. Performances in both experiments were measured in two conditions: (i) ball driven by a constant force field and (ii) by a variable force field. Results showed that while ball dynamics stability facilitates interception, it does not affect perceptual assessment. The dynamic visual information seems therefore to be processed differently when its purpose is a motor act versus a perceptual one.