Large-scale neuronal recordings have revealed that movement-related activity is distributed across the brain. Within motor cortices, such activity is constrained to latent dynamics on the population level. Whether other regions exhibit these motor cortical dynamics, and how this distributed activity relate to specific motor behavior remains unclear. We used multiple Neuropixels probes to simultaneously record population activity across cortical and subcortical regions during a goal-directed reaching task in mice. Task-related population activity followed dominant latent dynamics across all recorded regions and were preserved across recordings sessions and animals. The population activity predicted the reach onset and represented a time-varying signal that was continuously modulated by the movement amplitude and presence of reward. Furthermore, the distributed activity exhibited a temporal hierarchy, with earliest onset in thalamic and cortical regions. Together, our findings provide evidence for a global, continuous, and adaptive neural representation of actionable reward prediction allowing brain-wide computation based on momentary motor outcome.​