Copyright 2016 American Medical Association. All rights reserved. IMPORTANCE Double-seronegativemyasthenia gravis (dSNMG) includes patients with myasthenia gravis (MG) without detectable antibodies to the nicotinic acetylcholine receptor (AChR) or to muscle-specific tyrosine kinase (MuSK). The lack of a biomarker hinders the diagnosis and clinical management in these patients. Cortactin, a protein acting downstream from agrin/low-density lipoprotein receptor-related protein 4 (LRP4)/MuSK, has been described as an antigen in dSNMG. OBJECTIVE To describe the frequency and clinical features of patients with dSNMG who have cortactin antibodies. DESIGN, SETTING, AND PARTICIPANTS A retrospective cross-sectional studywas conducted at Hospital de la Santa Creu i Sant Pau, an institutional practice referral center in Barcelona, Spain, between May 1, 2015, and November 30, 2015.We included 250 patients with a definitive diagnosis of MG with available serum samples at the time of diagnosis. Descriptive and comparative data analyses were performed. EXPOSURES Cortactin antibodies were measured by enzyme-linked immunosorbent assay andWestern blot; AChR, MuSK, and anti-striated muscle antibodies were detected using a standard method; and LRP4 antibodies were tested using a cell-based assay. MAIN OUTCOMES AND MEASURES The primary outcomewas the frequency of patients with dSNMG who have cortactin antibodies. Secondary outcomes were demographic, clinical, neurophysiological, and laboratory data. RESULTS Of 250 patients (mean [SD] age at onset, 49.7 [21.2] years; 56%female), 38 (15.2%) had dSNMG, 201 (80.4%) had MG with AChR antibodies, and 11 (4.4%) had MG with MuSK antibodies. Cortactin antibodies were identified in 28 patients with MG: 9 of 38 (23.7%) who had dSNMG, 19 of 201 (9.5%) who had MG with AChR antibodies (significantly lower than those with dSNMG: 9.5%vs 23.7%; P = .02), and 0 of 11 who had MG with MuSK antibodies; 0 of 29 controls had cortactin antibodies. At onset, among the 9 patients with dSNMG and cortactin antibodies, 6 had ocular MG and 3 had Myasthenia Gravis Foundation of America clinical classification IIA. Two patients with ocular MG developed generalized MG. The group with dSNMG and cortactin antibodies, compared with those who had MG with AChR antibodies, more frequently had mild forms at onset (100.0%vs 62.7%; P = .03), had fewer bulbar signs at maximal worsening (0% vs 41.3%; P = .01), and were younger at onset (median [interquartile range], 34.9 [9.5] vs 53.9 [38.5] years; P = .03); the group with dSNMG and cortactin antibodies also more frequently had ocular MG at onset than those with MG and AChR antibodies, although the difference was not statistically significant (66.7%vs 40.8%; P = .17). Of 17 patients with ocular dSNMG, 4 (23.5%) had antibodies to cortactin. CONCLUSIONS AND RELEVANCE In this study, patients with cortactin antibodies and dSNMG had an ocular or mild generalized phenotype of MG. Including the detection of cortactin antibodies in the routine diagnosis of dSNMG may be helpful in ocular MG.