We normally view our own body from a 1st person perspective (1PP) whereas our view of other people is always from the 3rd person perspective (3PP). Thus, body parts viewed from the 1PP should be more readily associated with the self than when viewed from a 3PP. Using synchronous multisensory feedback, fake body parts, even multiple limbs, can be embodied when viewed from the 1PP (eg the rubber-hand illusion ö Botvinik and Cohen 1998; supernumerary limbs öNewport et al 2010). On the other hand, while viewing a body from a 3PP can induce an out-of-body experience (Ehrsson 2007), 3PP inhibits ownership of a generic fake body that is normally embodied from 1PP (Petkova et al 2011). Conversely, 3PP rubber hands are embodied when participants look at themselves in a mirror (Bertamini et al 2011) and both mirror viewing and 3PP video can restore self-awareness following stroke (Fotopoulou et al 2009, 2011). To investigate the extent to which real body representations can be updated through 3PP, participants wore a head-mounted display (EZ800) through which they viewed real-time (delay  33 ms) video images of themselves from a monochrome camera positioned 2 m behind them. Participants reclined to an almost supine position in an Optima TM wheelchair, such that the view through the camera was restricted to the top of their head and their outstretched arm raised towards the ceiling (figure1a). In each condition, the experimenter gripped the participant by the wrist and pulled vigorously upwards, providing congruent visuo-tactile feedback and stretching the tendons. During the stretch condition, the image of the arm was simultaneously elon- gated, stretching it to twice its normal length. The image remained unaltered in a non-stretch condition. Condition order and manipulated arm were counterbalanced between participants. Immediately before and after manipulation, the participants watched themselves as they waved their arm, as if casting a fishing rod, for 30 s, after which additional synchronous visuo-tactile feedback was applied by tapping along the length of the arm with a stick. As a final visuo-tactile `convincer' of arm length, a key was suspended out of reach above the hand prior to manipulation, remaining out of reach after non-stretch but being within reach after stretch (figure 1) whilst appearing to be at a constant height throughout. In the stretch condition, the convincer tasks were applied whilst the image remained elongated and lasted for 2 min. SHORT AND SWEET How long is your arm? Using multisensory illusions to modify body image from the third person perspective Perception, 2012, volume 41, pages 247 ^ 249 Catherine Preston Department of Neuroscience, Karolinska Institutet, Retzius va« g 8, 17177 Stockholm, Sweden Roger Newport½ School of Psychology, University Park, University of Nottingham, Nottingham NG7 2RD, UK; e-mail: roger.newport@nottingham.ac.uk Abstract. Updating body representations from the 3rd person perspectives (3PP) seems to require viewing the real body, unlike when viewing from a 1st person perspective. Here, 3PP updating was investigated through induction of a physically impossible multisensory illusion in which participants viewed real-time 3PP video of themselves having their arm pulled until it stretched to twice its normal length. The illusion elicited the subjective experience that the participant's own arm had been stretched and caused an overestimation of reaching distance, although actual reaches were unaffected. Multisensory illusions from the 3PP can alter body image when applied to real bodies. doi:10.1068/p7103 ½ Corresponding author.