Previous studies have suggested that the relative importance associated with redundant sensory signals should depends only on their relative noisiness, so as to optimize movement precision (Maximum Likelihood Principle, MLP). As a consequence, whether a goal-oriented hand movement is performed in a seated or standing position should not affect the visuo-proprioceptive sensory weights. To test this hypothesis, we asked healthy subjects to perform visuo-manual coordination tasks both seated and standing. Using virtual reality and quantifying the deviation of the subjects’ motor responses due to imperceptible sensory conflicts, we estimated the relative weight associated with visual information. In contrast to the predictions of MLP, preliminary results suggest that the visual weight significantly increases when standing. This experimental evidence advocates for an improvement of the current statistical models of multisensory integration; more precisely, the factors determining the relative importance of the sensory inputs should not be restrained to their reliability, but they should include additional elements such as the postural constraints.