PROJECT SUMMARY Accurate perception of the body’s orientation in the world is central to everyday life, as balance problems are common and serious, especially in older adults. Computational theory proposes that this ability requires the brain to learn and store an internal representation of gravity that can be used for perception and action, and research has shown that this model is based on visual, somatosensory, and vestibular signals. The proposed experiments investigate the contribution of the vestibular system to the generation of this internal model and test whether the same neural mechanisms provide the gravitational information used for perception and action. In Aim 1, macaques will be trained to visually discriminate the earth-vertical orientation while they are in random head/body tilt positions and then behaviorally tested after bilateral removal of the vestibular organs. If this manipulation abolishes the ability to perform this task at tilted positions, that will indicate that vestibular information is critical for the internal model of gravity. This aim will also examine how visual orientation discrimination adapts to vestibular receptor loss by recruiting extra-vestibular sensory cues, such as somatosensation, and whether active training is necessary for this compensation. Aim 2 will examine how gravity-dependent vestibular information affects motor function by measuring arm kinematics and muscle activity from agonist and antagonist muscles before and after bilateral removal of vestibular inputs. This aim will also measure adaptation after vestibular injury and the role of training in this compensation. Aim 3 will distinguish whether these gravity effects on perception and action are mediated by a shared thalamocortical pathway by probing for deficits in the tasks of Aims 1 and 2 during reversible inactivation of the anterior thalamus, where gravity-tuned cells have been reported. Taken together, these experiments are important for understanding the multisensory influences of gravity on perception and action, as well as the underlying neural circuits, and for revealing how motor learning can aid recovery from vestibular dysfunction.