Most individuals with myasthenia gravis (MG) complain of cognitive impairment, but empirical studies of cognition in MG have produced mixed results. In the present review, we critically examined the methodology and results of previous studies that investigated cognition in MG. Results from our review revealed that none of the studies met at least 50% of criteria under review. The most common shortcomings of previous studies included small sample size, no exclusion for visual difficulties in patients, inadequate assessment of mood, and poor control for prednisone use. Despite these methodological difficulties, mild impairments on measures of learning have been identified. These findings need to be replicated with adequate control of potential confounds before any conclusions can be made regarding cognition in this disease. Suggestions for design of future studies are provided.