The clinical impression that severe hyperreflexia is not consistently associated with severe spasticity in hemiplegia or hemiparesis due to vascular pathology was tested by objective measurement in 11 patients. The intensity of tendon reflex responses of the triceps and biceps brachii muscles was measured as the reflex sensitivity ratio (evoked twitch tension/stimulating hammer force). The degree of spasticity was measured as the threshold rate of elbow flexion or extension above which a myotatic response was elicited, determined by both a sudden increase in passive resistance and a burst of antagonist electromyographical activity. Rate thresholds were rank ordered for flexion data and for extension data with their corresponding reflex sensitivity ratios. Numerous individual discrepancies stood out. Correlation coefficients between rate thresholds and reflex sensitivities determined at two elbow angles were -0.52 and -0.35 during extension, and -0.47 and 0.03 during flexion. Differences in rate of stretch, differences in the sensory structures stimulated, and differences in the distortion of the organization of motor control may account for the difference in responses to the two tests.