OBJECTIVE: To evaluate the usefulness of the sural/superficial radial sensory nerve action potential amplitude ratio (SRAR) in the electrodiagnosis of diabetic polyneuropathy. METHOD: Nerve conduction study was performed in 80 diabetic patients and 31 normal adults. Standard nerve conduction studies were performed. Sural nerve response was recorded with the active electrode on the posterosuperior margin of the lateral malleolus and stimulation 14 cm proximally. Superficial radial nerve response was recorded with the active on the snuffbox and stimulation 10 cm proximally. Baseline-to-peak sensory nerve action potential(SNAP) amplitudes were measured, and SRAR were obtained. The diabetic group was subdivided into an electro physiologically normal group (DMNL, n=32) and neuro pathy group (DMPN, n=48). SRAR was compared among the control (NORM, n=31), DMNL and DMPN groups. RESULTS: SRAR was 0.344 in NORM, 0.314 in DMNL and 0.145 in DMPN, respectively. SRAR showed 77.1% sensitivity and 93.8% specificity. Sural SNAP amplitude sensitivity was 85.4% and specificity, 88.7%. CONCLUSION: The relatively high specificity of SRAR indicates its usefulness in the diagnosis of diabetic polyneu ropathy. However, as the superficial radial sensory SNAP amplitude decreased significantly in the diabetic groups compared to control, the SRAR is not superior to the sural SNAP amplitude in the diagnosis of diabetic polyneuropathy.