This article evaluates diagnostic sensitivity of minimal F-wave latency, sural/radial amplitude ratio (SRAR), dorsal sural/radial amplitude ratio (DSRAR), sympathetic skin response (SSR), and R-R interval variability (RRIV) for detecting early polyneuropathy in patients with glucose intolerance and diabetic patients. F-wave latencies were more prolonged in diabetic patients with normal and abnormal nerve conduction studies than control subjects (p < .001). SRAR was lower, SSR latency was more prolonged, and RRIV was lower in diabetic patients with abnormal nerve conduction studies than healty controls (p < .001). SSR latency was more prolonged and RRIV was lower in diabetic patients with normal nerve conduction studies than healty controls (p < .01, p < .05, respectively). DSRAR was lower in diabetic patients with normal and abnormal nerve conduction studies than control subjects (p < .001). DSRAR was also lower in patients with glucose intolerance than control subjects (p < .01). DSRAR was the most sensitive and specific test in either of diabetic patients with normal nerve conduction studies (sensitivity 66%, specificity 90%) and diabetic patients with abnormal nerve conduction studies (sensitivity 100%, specificity 90%). DSRAR is the most reliable method for detection of early nerve pathology. Patients with glucose intolerance might have subclinical neuropathy that can be demonstrated with DSRAR analysis.