Physicians are sometimes reluctant to refer patients for electrodiagnostic studies (electromyography with nerve conduction studies [EMG/NCS]) believing the test is too painful and of little benefit.We performed two separate surveys on 126 and 100 consecutive patients referred to our laboratory to determine if EMG/NCS was beneficial to the referring physician and to compare the level of anxiety experienced by patients before the study with the pain actually experienced during the study.The electrodiagnosis was discordant from the referring diagnosis in 39% of the patients with an abnormal EMG/NCS. Pretest anxiety levels were low in 59% of the patients, medium in 27%, and high in 14%. After the tests, 82% of the patients said that the test was not as bad as expected, and was generally only mildly painful. Ninety-three responded that they would have the test performed again.EMG/NCS often suggest alternative diagnoses, and the actual pain experienced during an EMG/NCS study is significantly less than expected.