This study examined the active joint-position sense in children with Spastic Hemiparetic Cerebral Palsy (SHCP) and the effect of static visual feedback and static mirror visual feedback, of the non-moving limb, on the joint-position sense. Participants were asked to match the position of one upper limb with that of the contralateral limb. The task was performed in three visual conditions: without visual feedback (no vision); with visual feedback of the non-moving limb (screen); and with visual feedback of the non-moving limb and its mirror reflection (mirror). In addition to the proprioceptive measure, a functional test [Quality of Upper Extremity Skills Test (QUEST)] was performed and the amount of spasticity was determined in order to examine their relation with proprioceptive ability. The accuracy of matching was significantly influenced by the distance that had to be covered by the matching limb; a larger distance resulted in a lower matching accuracy. Moreover it was demonstrated that static (mirror) visual feedback improved the matching accuracy. A clear relation between functionality, as measured by the QUEST, and active joint-position sense was not found. This might be explained by the availability of visual information during the performance of the QUEST. It is concluded that static visual feedback improves matching accuracy in children with SHCP and that the initial distance between the limbs is an influential factor which has to be taken into account when measuring joint-position sense. © 2011 Elsevier Ltd.