BACKGROUND: Previous studies of radial nerve conduction study (NCS) did not present how to measure the length of the radial nerve across the elbow, and did not even mention how to manage the spiral course of the nerve. This study aimed to applicate the most reliable method to measure the length of the radial nerve during NCS. METHODS: Three points (A, B, and C) were determined along the relatively straight course of the radial nerve. The distance was measured using three different methods: L1) straight distance corresponding to the A-C distance, L2) sum of the distances corresponding to the A-B-C distance, L3) based on the L2, but the elbow is flexed at a 45° angle. We compared the three methods of distance measurement and the calculated nerve conduction velocities (V1, V2, and V3) in normal healthy subjects. RESULTS: 19 normal participants were enrolled. The mean value for method L1, L2 and L3 were 22.5 ± 1.8 cm, 24.0 ± 2.1 cm, and 23.2 ± 2.1 cm (p < 0.001). Calculated conduction velocities using those distance measurement methods as follows (p < 0.001): V1 (60.9 ± 2.7 m/s), V2 (64.6 ± 3.3 m/s), and V3 (63.4 ± 3.9 m/s). V2 was significantly greater than V1 and V3 (p < 0.001, p = 0.010, respectively). CONCLUSIONS: The distance measurement using a stopover point near the lateral epicondyle between two stimulus points in position of a fully extended elbow with forearm pronation is the most appropriate posture for radial motor NCS.