ABSTRACT Humans can skillfully control their grasp during actions as complex and dynamic as swinging a tennis racket, and as simple and static as holding a briefcase. Both tasks require the use of sensory feedback to achieve and maintain an appropriate grasp force. There is evidence that motor and somatosensory cortices communicate task-relevant information in order to enable skillful movement. Our primary goal is to uncover the motor cortical dynamics underlying grasp force control and determine the extent to which these dynamics are mediated by behavioral context and corticocortical communication of somatosensory feedback. We propose to study the cortical control of grasp by leveraging the unique experimental paradigms afforded by a bidirectional human brain-computer interface study in which participants with tetraplegia have intracortical electrode arrays implanted in motor and somatosensory cortex. Previous work, primarily focused on reaching movements, has demonstrated that motor cortex exhibits population dynamics that are constrained within low- dimensional manifold. We have identified similar dynamic responses within human motor cortex that contain information about grasp force. However, these responses are task-dependent and can change as the complexity of the proximal arm movement changes. Here we will extend that work to study the context- dependence of M1 dynamics across a range of static and dynamic hand and arm movements including both overt and covert (i.e., imagined) behaviors. Sophisticated motor control relies on sensory information to shape neural control signals emanating from motor cortex, yet very little is known about the flow of information from somatosensory to motor cortex for the control of the hand. We aim to quantify the corticocortical communication pathways across a range of task contexts through the analysis of simultaneous neural recordings in motor and somatosensory cortex. We will then use intracortical microstimulation to probe these communication pathways while providing task-relevant sensory feedback as well as task-irrelevant stimulation as a control. Finally, we will use a brain-computer interface to test whether there is the potential for plasticity within the corticocortical communication circuits, or whether communication is constrained by between-area dynamics. Successful completion of this proposal will lead to new knowledge about the role of M1 in dynamic and static grasp behaviors. We will quantify how somatosensory input is communicated with M1 and whether corticocortical communication pathways can be modified through training, which has relevance to understanding skill learning and improving rehabilitation.