Clinical findings have limited value in predicting electrophysiologically confirmed median neuropathy at the wrist (MNW). To determine the value of clinical findings and an automated electrophysiologic neurodiagnostic device (AEND) in diagnosing MNW, we studied two groups of 75 consecutive patients (an initial group and a validation group, 150 total) referred to an academic electrophysiology laboratory for upper extremity complaints. The definitive standard for MNW was the neurologist's diagnosis after formal clinical and electrodiagnostic evaluation. The neurologist was blinded to the results of the AEND (NC-Stat, NeuroMetrix, Inc). In the validation group, the AEND yielded a distal motor latency (DML) in 97% of hands with a conventional motor response, and the correlation of the AEND DML with the conventional DML was 0.94 (P < 0.001). Of 248 symptomatic hands, the neurologist diagnosed 117 (47%) with MNW. At 90% specificity, the AEND DML had a sensitivity of 86% for MNW. Age, body mass index, sensory symptoms in digits 1 to 3, and nocturnal awakening were independent clinical predictors of MNW. Each 1-msec increase in the adjusted AEND DML was independently associated with an OR of 298 (95% confidence interval, 40 to 2233) for MNW. Each 1-msec increase in the F-wave latency was independently associated with an OR of 2.6 (95% confidence interval, 1.3 to 4.9) for MNW. Compared with a model based solely on clinical variables, an algorithm including symptom variables plus the AEND DML had an odds ratio for correct diagnostic classification of 6.3 (95% confidence interval, 3.8 to 12.3). The sensitivity at 90% specificity improved from 40% for the clinical model to 86% for the model with DML. A practical method for integrating clinical and electrophysiologic findings to assess the risk of MNW was proposed. This method correctly stratified 79% of control and MNW patients into very low- and high-risk groups, respectively. We concluded that MNW diagnosis is significantly improved with an AEND.