A coordinated triggering task requiring use of kinesthetic information was employed to assess the timing of use of kinesthetic information in normal subjects and patients with cerebellar dysfunction. Passive movements of varying velocity were imposed in the flexor direction about the metacarpophalangeal joint of the right index finger. Subjects attempted to depress a switch with their left thumb when the index finger moved, past a specified angle that was learned during a training session. The velocities ranged from 10°/s to 88°/s in 2°/s increments. After 200 trials, subjects were then instructed instead to react as quickly as possible (reaction-time task) to the onset of movement for an additional 200 trials. For the same movements, the timing of onset of responses of muscle spindle afferents and cutaneous mechanoreceptors was determined by recording the responses of these afferents using microneurography. For slow velocities, patients were able to perform similarly to normals but at faster velocities patients triggered too late compared with normals. Patients required more time to use kinesthetic information than did normal subjects. An estimate of kinesthetic processing was not longer in patients. The chief explanation for the prolonged time required to use kinesthetic information in patients was that their reaction times were prolonged by 93 ms. In addition, the movement time was also prolonged, but this accounted for only 23 ms. Impaired motor performance in tasks requiring the use of kinesthetic information in cerebellar patients can be explained largely by their prolonged reaction times. Muscle spindle afferents responded on average much sooner than cutaneous mechanoreceptors. Because of the limited time available to perfomr the kinesthetic triggering task, the role for cutaneous mechanoreceptors, to provide singals for on-line coordination of movement appears limited compared with muscle spindle afferents.