Background and Purpose-The active force generated by a single muscle fiber is greatest in midrange. In healthy subjects, the reduced muscle force at short muscle length is partially compensated by modified patterns of muscle activation, probably central in origin. These patterns are presumed to be deficient after stroke. We examined the excitability muscle-length relation in stroke patients and healthy controls and hypothesized about its alteration in stroke patients. Methods-Corticospinal excitability was assessed in 31 stroke patients (19 subcortical, 12 cortical) and 19 healthy volunteers by transcranial magnetic stimulation. We recorded the motor evoked potentials (MEPs) simultaneously from the biceps brachii and the triceps brachii muscles at 0°, 20°, 40°, 60°, 80°, 100°, and 120°degrees of elbow flexion (0°b eing full elbow extension). Results-Normal subjects revealed a significant increase in MEP amplitudes at shortened muscle lengths for both the flexor and extensor muscles (PϽ0.001). Multivariate variance analysis revealed that the MEP-angle curves of cortical stroke patients were significantly different from those of the control group for both muscles, lacking an increase of corticospinal excitability at short muscle length. Yet the MEP-angle curves for the subcortical stroke patients did not show a statistically significant difference from the control group for either muscle. Conclusion-Cortical and subcortical strokes differentially affect the corticospinal excitability muscle-length relation. This may account for the reported disproportionate decrease in muscle strength at shortened range after stroke.