Nerve conduction studies (NCS) are one of the basic tools of the electrodiagnostic clinician and are performed in order to evaluate the integrity of peripheral nerve function . Using such techniques diagnosis of various diseases and disorders of nerve are possible. During NCS a peripheral nerve is stimulated using an electrical pulse of sufficient intensity to recruit as many nerve fibres as possible. This elicits a volley of action potentials (APs) in the individual nerve fibres which is then recorded at a distal point usually using surface electrodes. Usually only two components of the recorded response (called the compound action potential or CAP), the distal amplitude and the distal/peak latency are routinely recorded though the proximal latency is also used to construct an average measure of conduction velocity (CV). In this study a model of the sensory CAP of the median nerve as measured using bipolar electrodes is proposed and using this model an additional measure is extracted from the CAP. This measure represents the distribution of conduction velocities within the nerve trunk and from this a measure of nerve fibre diameter distribution can be ascertained