Subsequent to prism adaptation, subjects may be readapted to their original visual-motor coordination using procedures similar to those occurring during adaptation. It has been argued that such procedures will be more effective than allowing for a decay of the aftereffect if normal visual-motor behavior is itself astate of adaptation. In the present study, two readapta tion procedures were compared for their effectiveness to four decay and/or control conditions. Although all groups, except for one decay condition, showed significant reductions in prism aftereffects, only the two readaptation groups showed final aftereffects that were not signif icantly different from zero. It is argued that readaptation is distinguishable from decay and, further, that the necessary conditions for readaptation can be elucidated with reference to the information discordance hypothesis of prism adaptation. These arguments are primarily based on an extensive analysis in which discrepant information about the adapted arm's position is supplied by a nonvisual search task using the two arms.