Sir Francis Walshe, in discussing "The Future of Neurology" at a meeting of the Royal Society of Medicine, stated his conviction that "a fruitful future for neurology lay rather with experimental pathology than with electrophysiological studies. In the case of the latter, it seemed that many studies revealed more about the properties of the apparatus used than of the functions of the nervous system."1 Thus, whenever a relatively new technique involving complicated apparatus, such as ocular electromyography, is introduced, one must establish normal ranges for proper interpretation of the results. More important, one must be able to distinguish artifacts from meaningful data. A previous publication2 outlines the technique of human electromyography. The purpose of this paper is to point out (A), normal variations of ocular electromyography, and (B), artifacts in electromyography. Because of these normal variations and artifacts, we shall, in our discussion, point out some limitations