SUMMARY In 143 normal subjects latencies of H-reflexes of the flexor carpi radialis muscle were well correlated with arm lengths. In 32 patients with injury to cervical roots C5 to C8 it was shown that latencies were only abnormal in lesions of roots C6 or C7. It is suggested that in most instances root C6 or C7 and, probably both roots in some cases are involved in the reflex pathway. Repeat electrophysiological examinations after surgery provided valuable data about the various types of nerve root recovery.Although the H-reflex of the flexor carpi radialis muscle is easily obtainable its diagnostic value has so far been scantily evaluated.'-3 The cervical nerve root involved in this reflex is unclear.In a previous study on changes in the flexor carpi radialis H-reflex in radiation-induced brachial plexus lesions, parameters have been determined for latency and conduction velocity.3 In this report normal latencies of the H-reflex were correlated with arm lengths. In addition, H-reflexes in patients with compression of a single cervical root shown by myelography were studied in order to gain information regarding H-reflex changes in such lesions and to evaluate the nerve root involved in the reflex pathway. Methods and materialsFlexor carpi radialis EMG and H-reflexes in both arms were performed by needle electrodes in both controls and patients. The method used to investigate the flexor carpi radialis H-reflex has been described previously.3 The median nerve in the cubital fossa was stimulated using a bipolar surface electrode with the cathode proximal. The stimulus frequency was one per five seconds, and the pulse duration 05 ms. The voltage was increased until a maximal H-reflex amplitude was obtained. After that supramaximal shocks were delivered to produce the maximal direct muscle (M) response. The H-reflex latency (H-RL) was measured at the beginning of the maximal H-reflex and the interlatency time (ILT) calculated by subtracting the max-