Abstract: To survive in a rapidly dynamic world, a key principle of cognitive and neural processing is to predict the future state of the world, and to proactively adjust perception and action. It is becoming recognized that this takes place not only for the “where” and “what” dimensions, but also the “when”, based on temporal regularities in the environment. I will present studies in which I investigated the cognitive architecture and neural basis of temporal anticipation. Using computational modelling and EEG, I will identify overlapping and distinct mechanisms of temporal anticipation when relying on interval memory or rhythm synchronization, challenging dominant oscillation-based theories. Next, taking a neuropsychological approach, I will present data from subcortical degeneration patients pointing to a double dissociation between rhythm- and interval-based prediction in the cerebellum and basal ganglia, respectively. Combining the two approaches, I will show that the cerebellum controls the temporal adjustment of cortico-striatal EEG neural dynamics, and use computational models to identify cerebellar-controlled computation. Altogether, these findings reveal that cognitively, temporal anticipation operates through distinct context-specific computations, but that lead to similar modulations both at attentional and motor preparation stages. Neurally, these findings establish the role of subcortical structures in temporal anticipation, revising the dominant cortico-centric view of dynamic cognition.