Six typical patients with the carpal tunnel syndrome in a neurological practice were double blindly treated with pyridoxine and placebo to determine whether electromyographic data alone could identify without mistake those patients receiving pyridoxine or placebo. All six patients had a significant deficiency of vitamin B6 by the data on the erythrocyte glutamic oxaloacetic transaminase. Multiple measurements for statistical significance of EMG functions were made on temperature-controlled hands over periods of 15 to 27 days for control and final evaluation. There were significant changes in criteria of amplitude on placebo which ruled out these criteria for a double blind trial. There was no statistically significant change in any criterion of latency and conduction velocity for patients on placebo, indicating these criteria could serve for differentiation. Criteria of latency and conduction velocity showed significant changes for three patients on pyridoxine, and the extent of changes indicated a relationship to compliance. By data on latency and conduction velocity, no mistake was made in blindly identifying patients receiving pyridoxine or placebo; p less than 0.0156.