PROJECT SUMMARY The cerebellum is a brain structure long known to be essential for coordinating the contraction of muscle groups across joints to enable smooth and precise limb movement. Output pathways in the cerebellar nuclei are thought to continuously generate rapid corrective signals that ensure precision during skilled movements through a process termed online correction. Up until now it has been difficult to identify and characterize the specific neural circuits that could implement this rapid refinement due to the lack of selective access to cerebellar output pathways. Recent work has shown that a subset of cerebellar output neurons project directly to the spinal cord (cerebello-spinal), providing a possible pathway for rapid and direct adjustments of the limb. However, little is known about the precise influence direct cerebello-spinal (CbSp) projections have on motor output and the timescale on which activity in these circuits may act to ensure the accuracy of dexterous behaviors. No research has explicitly investigated whether direct projections from the cerebellum to the spinal cord mediate rapid, online corrections. The overarching goal of this proposal is to define how output from the cerebellum enables skilled movements, focusing specifically on the functional role of CbSp projections and their influence on forelimb movements. The central hypotheses are: 1) CbSp projection neurons convey online corrective commands that refine forelimb movement; and 2) this refinement is achieved through activity patterns that encode predictions about limb kinematics or muscle activity. Employing a skilled water reaching assay, CbSp projections will be optogenetically silenced during performance of behavioral tasks designed to introduce sources of movement error, such as changing reach target location. Kinematic and electromyography (EMG) analyses of performance will uncover the precise corrective role of CbSp neurons in forelimb movements. Next, multielectrode silicon probes will be used to record from CbSp neurons during performance of the same water reaching tasks. Single unit activity analyses and generalized linear models trained on kinematic and neural activity data will reveal whether CpSp activity predicts corrective movements and encodes specific features such as muscle recruitment, limb velocity, acceleration, or trajectory. Moreover, analysis of data from both Aims will determine whether CbSp neurons mediate corrective signals to the forelimb during dexterous movements. This work will provide valuable insight into the neural basis of dexterous movement by expanding knowledge of how the cerebellum facilitates the speed and precision of forelimb behaviors. This research will help lay the groundwork for improved diagnosis and treatment of cerebellar pathologies.