The role of proprioception in the control and adaptation of visuomotor rela
tionships is still unclear. We have studied a deafferented subject, IW, and
control subjects in a task in which they used single joint elbow extension
to move to a visual target, with visual feedback of the terminal position
provided by a cursor displayed in the plane of their movements. We report t
he differences in movement accuracy between the deafferented subject and co
ntrols in the normal task and when challenged with a cognitive load, counti
ng backwards. All subjects were less accurate when counting; this was a sma
ll effect for the controls (<10% change) but much greater for the deafferen
ted subject (>60% change). We also examined changes in movement kinematics
when the instructed amplitude was altered via a changed gain between final
arm position and presentation of the feedback cursor. The deafferented subj
ect maintained temporal movement parameters stable and altered amplitude by
scaling force (i.e. changed peak velocity), whereas the controls scaled bo
th movement velocity and duration. Finally, we compared the subjects' adapt
ation of movement amplitude after a period of exposure to the changed visuo
motor gain. The deafferented subject was able to adapt, but his adaptation
was severely impaired by the counting task. These results suggest that prop
rioception is not an absolute requirement for adaptation to occur. Instead,
proprioception has a more subtle role to play in the adjustment to visuomo
tor perturbations. It has an important role in the control of reaching move
ments, while in the absence of proprioception, attention appears necessary
to monitor movements.