We investigated whether the perception or production of a given line length in normal subjects varies according to where in peripersonal space the line is perceived or produced. We also investigated the influence of the direction of movement used to make the line. In Experiment 1, blindfolded normal subjects were asked to estimate distances while the examiner moved the subject's hand in proximal (medial) or distal (lateral) space, moving centripetally or centrifugally. The subjects showed a spatial effect, perceiving the same length as shorter in proximal space than distal space. This result could be related to either a proximal spatial attentional bias or an anisometric representation of spatial distances. In Experiment 2, we attempted to dissociate these hypotheses by studying blindfolded normal subjects, who were requested to produce horizontal lines of a given length (100 or 200 mm) in proximal versus distal peripersonal space using centripetal or centrifugal movements. Centrifugal movements in proximal space were the longest; centrifugal movements in distal space were the shortest; in between were the proximal centripetal and distal centripetal movements which did not differ from each other. These results suggest that in peripersonal space the perception of length in normal subjects is most consistent with anisometric mental representation where the size of mental representations of length units decreases as a function of the distance from the subject's midsagittal plane. Length production, however, may depend on an interaction of the anisometric mental representation and the premotor/intentional factors. (JINS, 2004, 10, 913–919.)