To compare the sensitivities for detecting cognitive impairment in patients with multiple sclerosis (MS) and administration times of three brief batteries of neuropsychological tests, 64 patients with MS completed the Neuropsychological Screening Battery for Multiple Sclerosis (NPSBMS), the Screening Examination for Cognitive Impairment (SEFCI), and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Failure on a particular test was defined as a score below the 5th percentile for healthy controls, and the number of patients who failed at least one or two tests (out of four) was determined for each battery. Both the SEFCI and the NPSBMS identified significantly more patients with impairment than the RBANS, which was no more sensitive than the Mini-Mental State Exam (MMSE). Results were similar at both the one- and two-failed-tests criteria, but there were no significant differences between the SEFCI and the NPSBMS at either failure criterion. Mean administration time was 22.6 min for the SEFCI compared to 31.7 min for the NPSBMS (p < 0.001). Eleven (17%) of the patients refused to attempt the Paced Auditory Serial Addition Test (PASAT), one component of the NPSBMS. For screening patents on a single occasion, the SEFCI is preferred because its administration time is shorter than the NPSBMS.