Objective: To investigate the validity of automated nerve conduction studies compared to traditional electrodiagnostic studies (EDS) for testing median nerve abnormalities in a working population. Design: Agreement study and sensitivity investigation from 2 devices. Setting: Field research testing laboratory. Participants: Active workers from several industries participating in a longitudinal study of carpal tunnel syndrome. Methods: Sixty-two subjects received bilateral median and ulnar nerve conduction testing across the wrist with a traditional device and the NC-stat automated device. We compared the intermethod agreement of analogous measurements. Main outcome measurement: Nerve conduction study parameters. Results: Median motor and sensory latency comparisons showed excellent agreement (intraclass correlation coefficients 0.85 and 0.80, respectively). Areas under the receiver operating characteristic curves were 0.97 and 0.96, respectively, using the optimal thresholds of 4.4-millisecond median motor latency (sensitivity 100%, specificity 86%) and 3.9-millisecond median sensory latency (sensitivity 100%, specificity 87%). Ulnar nerve testing results were less favorable. Conclusion: The automated NC-stat device showed excellent agreement with traditional EDS for detecting median nerve conduction abnormalities in a general population of workers, suggesting that this automated nerve conduction device can be used to ascertain research case definitions of carpal tunnel syndrome in population health studies. Further study is needed to determine optimal thresholds for defining median conduction abnormalities in populations that are not seeking clinical care.