Cerebellar-cortical beta oscillations emerge as a predictive signal facilitating the stability of behavioral performance.

Adaptive behavior enables individuals to respond flexibly to environmental changes by forming expectations based on experience within the new environment. Beta oscillations (13-30 Hz), with their widespread distribution,1,2,3,4,5,6,7,8,9,10,11,12 play a central role in this process.13,14,15,16,17,18,19,20 Specifically, beta synchronization occurring 2 s before movement initiation is modulated by prior errors21,22,23 and may reflect predictions based on past outcomes.24,25,26,27 Yet, the spatiotemporal dynamics of pre-movement beta oscillations, as well as their roles in detecting environmental changes and in iteratively updating motor plans to optimize and stabilize performance, remain elusive. Here, we reveal that beta oscillations emerge in a cerebello-cortical network 2 s before action initiation and progressively build up across trials as environmental features are learned and behavioral outcomes become more stable. Within this network, directional connectivity analyses reveal that the cerebellum initially drives prefrontal activity during the pre-movement period, with this influence reversing near movement onset. Finally, using a single-trial approach, we establish that, before action initiation, beta bursts in this network predict performance in the upcoming trial based on previous outcomes. These findings identify pre-movement beta oscillations within a cerebello-cortical network as a neural substrate supporting predictive processes that stabilize motor performance across changing environments. They emphasize the contribution of cerebellar networks to cognitive aspects of motor control up to 2 s before movement onset.