Since Sherrington's convincing demonstration of the reciprocal innervation of opposing muscles, it has generally been thought that antagonist muscles are inactive during most voluntary movements. However, more recent evidence suggests that excitation of Renshaw cells may facilitate antagonist coactivation whereas excitation of Ia inhibitory neurons can induce reciprocal inhibition. A body of evidence has accumulated to indicate some of the circumstances which particularly favour the co-contraction of antagonist muscles. Isometric prehension, either in the precision grip or the power grip, can be shown to be one of the most important examples of antagonist coactivation. Studies of the discharge of single Purkinje cells of the intermediate cerebellar cortex in awake monkeys during performance of a maintained grip revealed that the majority of these neurons are deactivated during antagonist co-contraction. In contrast, other, unidentified neurons of the cerebellar cortex were as a group activated during grasping. It is suggested that the Purkinje cells act to inhibit antagonist muscles during reciprocal inhibition but are themselves inhibited during antagonist coactivation. These results support a suggestion made by Tilney and Pike in 1925 that the cerebellum plays an important role in switching between the coactivation and reciprocal inhibition of antagonist muscles.