Since there is a need for cost-effective screening techniques to identify neuropsychological impairment in multiple sclerosis (MS) patients, and because existing methods require cognitive testing with subsequent interpretation by a neuropsychologist, a brief self-report procedure was developed to screen for neuropsychological impairment in MS. In the first phase of the study, a pool of 80 items was generated based on a literature review and consultation with healthcare professionals. The set was reduced to 15 via Rasch analysis. Using these items, a brief (five minute) MS Neuropsychological Screening Questionnaire (MSNQ), including patient- and informant-report forms, was composed. In phase II, 50 MS patients and their caregivers completed the MSNQ. A comprehensive neuropsychological test battery was also administered. Analyses covered the reliability of the MSNQ and correlations between both patient- and informant-report scores and objective neuropsychological testing. Cronbach's alpha coefficients were 0.93 and 0.94 for the patient- and informant-report forms, respectively, and both forms of the test were strongly correlated with a more general cognitive complaints questionnaire. The patient MSNQ form correlated significantly with measures of depression but not with objective tests of cognitive function. In contrast, the informant form was correlated with patient cognitive performance but not depression. A cut-off score of 27 on the informant form of the MSNQ optimally separated patients based on a neuropsychological summary score encompassing measures of processing speed and memory. There were two false-negatives and one false-positive, giving the test a sensitivity of 0.83 and a specificity of 0.97. It is concluded, therefore, that this self-administered neuropsychological screening test is reliable and predicts neuropsychological impairment in MS patients with a reasonable degree of accuracy.