Hyperactive strethch reflexes in the upper motor neuron (UMN) syndrome are frequently cited as an impediment to volitional movement. The assumption is that neural or mechanical activity of the hyperactive antagonist interferes with agonist function. The validity of this assumption was examined by evaluating quantitative and qualitative relationships between stretch reflexes and voluntary movement. Sixteen patients with chronic UMN symptoms and 8 normal volunteers were tested. Joint position and integrated electromyograms from primary flexors and extensors were recorded. Quantitated values of (1) reflex response to controlled passive motion by an automated system, (2) a maximal voluntary isometric contraction, and (3) the time required for ten voluntary rapid repetitive movements (RRM) of alternating elbow flexion and extension were obtained. Passive movement elicited tonic reflexes, which predominated during muscle stretch in patients and during muscle shortening in the volunteers. Ratios of the EMG activity elicited during stretch, shortening, and isometric activity were used as measures of spasticity and were compared with the time for RRM. A positive correlation between elbow flexor spasticity and the time for RRM was shown. Qualitative analysis of the EMG activity during voluntary isotonic movement, however, showed that primary impairment of movement is not due to antagonist stretch reflexes, but rather to limited and prolonged recruitment of agonist contraction and delayed cessation of agonist contraction at the termination of movement.