Myasthenia gravis is an organ-specific autoimmune disorder generally thought to be caused by an antibody-mediated attack against the skeletal muscle nicotinic acetylcholine (Ach) receptor (AchR) at the neuromuscular junction. Extraocular muscle weakness and double vision are present in about 90% of patients with myasthenia gravis and are the predominant complaints in about 20% of patients, when the condition is called ocular myasthenia gravis (OMG). While serum antibodies against the AchR are detected in most patients with generalized myasthenia gravis (GMG), they are not found in about one-third of patients with the ocular variety, and epidemiological, clinical, and serological studies suggest that OMG and GMG are two separate diseases. Both forms of myasthenia gravis are sometimes associated with thyroid autoimmunity or thyroid-associated ophthalmopathy (TAO). We have therefore tested the sera of patients with GMG and OMG by Western blotting for antibodies against porcine eye muscle membrane proteins in general, and by enzyme-linked immunosorbent assays (ELISA) specifically for reaction with two skeletal muscle antigens which are prominent marker antigens for TAO, namely, the calcium-binding protein calsequestrin and the so-called "64-kDa protein." The 64-kDa protein has recently been identified as the flavoprotein subunit of mitochondrial succinate dehydrogenase. Patients with ophthalmopathy and myasthenia were excluded. Nine of the patients had associated Graves' hyperthyroidism without evident ophthalmopathy and one had Hashimoto's thyroiditis. Antibodies against porcine eye muscle membrane antigens of M(r) 15-110 kDa were detected in patients with GMG or OMG, one or more antibodies being detected in 100% of patients with GMG and in 88% of those with OMG. The most frequently found antibodies were those targeting eye muscle membrane proteins of 15, 67, and 110 kDa. Antibodies reactive with purified calsequestrin (63 kDa) were detected in 21% of patients with OMG but in no patient with GMG. Antibodies recognizing purified succinate dehydrogenase (67 kDa) were found in 42% of patients with OMG, in 100% (5 of 5) of patients with GMG, and in 48% of all patients with myasthenia gravis not associated with Graves' hyperthyroidism. There was no close correlation between any eye muscle-reactive antibody and antibodies against the AchR in either group of myasthenic patients. The findings support the notion that immunoreactivity against skeletal muscle proteins other than the AchR may play a role in the development of the muscle weakness in AchR antibody-negative patients with OMG and GMG, although it is unlikely that any of the antibodies demonstrated in this study are directly implicated. Similarly, while the demonstration of antibodies reactive with eye muscle antigens associated with TAO in patients with OMG raises the possibility that the link between the ocular lesions of myasthenia gravis and Graves' disease may be autoimmunity against a common antigen(s), it is more likely that both disorders are mediated by cytotoxic T cells recognizing another cell membrane antigen, such as the novel thyroid and eye muscle shared protein G2s, and that serum antibodies reactive with succinate dehydrogenase Fp subunit and calsequestrin are markers of an immune-mediated eye muscle reaction.