The MS Functional Composite (MSFC) is a continuous scale of neurological disability for patients with multiple sclerosis (MS). Cognition is represented by the Paced Auditory Serial Addition Test (PASAT), although the Symbol Digit Modalities Test (SDMT) has been proposed as a promising alternative. MSFC scores were calculated using either the PASAT or the SDMT with the following reference populations: National Multiple Sclerosis Society (NMSS) Task Force, 400 MS patients, and 100 normal controls. A subgroup of 115 patients was followed longitudinally, with a test-retest interval of 2.3 +/- 1.2 years. Pearson correlations were calculated and analyses of variance (ANOVAs) were used to assess relationships among the MSFC components and composite scores, and differences in performance between patients and controls. Longitudinal changes were also assessed. Logistic regression was performed to determine which MSFC scores are most predictive of diagnosis, course, and work disability. All MSFCs had similar test-retest reliability and correlations with other measures including neurological disability, depression, and fatigue. The SDMT showed slightly better validity with respect to predicting diagnosis, course, and work disability, although the amount of variance accounted for was similar for each version of the MSFC. Our data, derived from a large sample of MS patients and normal controls, supports the validity of both PASAT and SDMT versions of the MSFC. Because the SDMT has slightly better predictive validity and has a relatively easier administration procedure, some clinicians and researchers may wish to replace the PASAT with the SDMT in future calculations of the MSFC using the calculation methods provided in this manuscript.