Objective
There is a room for considerable error in the measurement of across-elbow conduction velocity due to the different possible positions of the elbow and the difficulty in measuring distance accurately. We propose a technique for the measurement of conduction velocity through the elbow segment in a fully flexed elbow position with the arm abducted at 90^o.

Method
We assumed 'ideal' across-elbow segmental conduction velocity is the mean of the forearm and arm segmental conduction velocities, and established an optimal deflection point at the elbow, which best reflects the ideal conduction velocity in normal healthy subjects. Five deflection points were examined at the elbow. Segmental conduction velocities of across-elbow segments were calculated at each of these points, using the sum of the linear distances from each point to the proximal above-elbow cathode stimulation site and to the distal below-elbow cathode stimulation site.

Results
The optimal deflection point was the midpoint between the epicondyle and the olecranon in an arm abducted 90^o and elbow fully flexed position.

Conclusion
Our data suggests that an across-elbow segment velocity lower than 54.2 m/sec, or a difference of more than 11.6 m/sec between the across-elbow and forearm segments is to be considered abnormal. The lower limit values expressed as mean - 2 S.D. for absolute across-elbow segmental conduction velocity and relative velocity difference between the across- elbow segment and forearm segments at other possible deflection points of the elbow were also calculated.