Clinical registries may facilitate research on myasthenia gravis (MG) in several ways: as a source of demographic, clinical, biological, and immunological data on large numbers of patients with this rare disease; as a source of referrals for clinical trials; and by allowing rapid identification of MG patients with specific features. Physician-derived registries have the added advantage of incorporating diagnostic and treatment data that may allow comparison of outcomes from different therapeutic approaches, which can be supplemented with patient self-reported data. We report the demographic analysis of MG patients in two large physician-derived registries, the Duke MG Patient Registry, at the Duke University Medical Center, and the INNCB MG Registry, at the Istituto Neurologico Carlo Besta, as a preliminary study to assess the consistency of the two data sets. These registries share a common structure, with an inner core of common data elements (CDE) that facilitate data analysis. The CDEs are concordant with the MG-specific CDEs developed under the National Institute of Neurological Disorders and Stroke Common Data Elements Project. © 2012 New York Academy of Sciences.