The spinal stretch reflex, exemplified by the tendon jerk, appears to be less important in humans than a delayed 'long-latency' response. This is easily observed when muscles of the hand are stretched while they are already contracting voluntarily. On limited evidence, many have long held that the delayed response is a transcortical reflex and have tended to neglect alternative possibilities, particularly that it might be a spinal reflex dependent upon slow afferents. New experiments have now eliminated the alternatives, leaving the transcortical hypothesis in command of the field.