Coordinated cyclical movements of ipsilateral elbow and knee or wrist and ankle can be performed more stably when the joints rotate in the iso-direction (e.g. wrist extension coupled to ankle dorsiflexion when forearm is in the prone position) than in the opposite-direction. This holds true regardless of the coupling of actions (e.g. extension-dorsiflexion, extension-plantarflexion), since the same results are obtained whether the forearm position is prone or supine. We investigated whether this direction dependence of the stability of coordinated movements was applicable to the coordinated movements of fingers and toes. Subjects (eyes closed) performed cyclical extension-flexion of fingers and toes in the two direction modes: iso- or opposite-direction, and the two coupling of actions modes: simultaneous (e.g. fingers flexion coupled to toes flexion) or alternate actions (e.g. fingers flexion coupled to toes extension). The stability of the movements was affected by the direction and the coupling of actions independently. In an additional experiment, motor evoked potentials (MEPs) of finger flexor and extensor muscles were elicited while subjects performed cyclical toes extension-flexion movements without activities of hand muscles with forearm prone or supine position. MEPs of finger extensors and flexors were modulated depend on the coupling of actions, not on the direction. In conclusion, the coordinated movements of fingers and toes were affected by both the direction and the coupling of actions, and the latter was supported by the neural coupling of fingers and toes. [J Physiol Sci. 2008;58 Suppl:S193]