Background: People with multiple sclerosis (MS) commonly experience cognitive impairment associated with the disease, but there is currently no agreed-upon operational definition for identifying the presence of that impairment, in either research or clinic contexts. The International MS Cognition Society (IMSCOGS) established a task force to begin to examine this issue and this paper represents the results of an initial pilot investigation. The aim of this paper was to compare two criterion sets to determine how to identify cognitive impairment among people with MS: the general Diagnostic and Statistical Manual (DSM-5) Criteria for neurocognitive disorders and criteria derived from existing MS research (scores in two domains fall 1.5 standard deviations below normative controls). Methods: Two hundred and ten people with MS presented for a brief cognitive evaluation in an MS Multidisciplinary Clinic at a midwestern academic medical center in the United States. Participants were generally middle aged (average 51.5 years), female (73.8%), and white (93.3%). McNemar's test was computed to compare the number of individuals whose cognitive test score performance was deemed cognitively normal, mildly impaired, or more significantly impaired. Results: DSM-5 criteria classified 87.2% of the sample as cognitively impaired, where 66.7% were more mildly impaired and 20.5% more significantly impaired. By contrast, research-based criteria classified 63.3% of the sample as cognitively impaired, with 49.5% as mildly impaired and 13.8% as more significantly impaired. Conclusions: These findings indicate that compared to research criteria, the DSM-5 criteria classified far more people with MS as having cognitive impairment secondary to the disease. The paper discusses the potential benefits and drawbacks of the two diagnostic methods, highlighting that more work will be needed in order to establish a standardized and validated method for characterizing these impairments.