This project will investigate children's capacity to separately adjust their neuromotor control to account for their own rapidly changing bodies and for forces imposed by novel external objects. There is growing evidence indicating that humans learn distinct predictive strategies for moving their own arm and for moving external objects, and humans use these strategies in an appropriate self/object context. Researchers have recently shown that this context-specific self/object distinction within the motor system is a developmental achievement that appears at about seven years of age in typically developing children. In addition, preliminary results from a study of children with high functioning autism (HFA) indicate that they are as capable as typically developing children of learning to compensate for a perturbing robotic manipulandum (object) but are delayed in learning not to carry this over to the free reaching (self) context. The overall aims of this project are to confirm this preliminary finding in children with HFA and to test a hypothesis about how to accelerate self/object context-specific dynamic motor learning in typically developing children as well as in children with HFA. The outcomes of this project will provide a better understanding of the development of context-specific dynamic motor adaptation in typically developing children, a clearer understanding of motor deficits in autism, and a therapeutic method for enhancing motor learning in children with HFA - especially learning to interact with dynamic objects. A positive outcome will also open future avenues for investigating typical and disordered development of motor interactions between individuals. The outcomes will be a better understanding of motor development in typically developing children, a clearer understanding of motor deficits in autism and a potential therapy for improving the development of dynamic motor interactions.