Objective: To establish reliability and validity of the Guy's Neurological Disability Scale (GNDS) for the assessment of patients with multiple sclerosis (MS) and to investigate whether it can be used by postal questionnaire.

Design: An observational study assessing one group of 22 patients using the GNDS face-to-face, repeating the assessment two or three weeks later with other tests; and assessing a second group of 21 patients first using a postal version of the GNDS and then seeing them face-to-face.

Setting: Patients in the community attending a day centre or a voluntary support group.

Subjects: Patients with multiple sclerosis.

Measures: The GNDS was the primary measure, and was compared with the Barthel Index, the Expanded Disability Status Scale (EDSS), the Short Orientation Concentration and Memory Test, the Nine-hole Peg Test, the Rivermead Mobility Index, the Hospital Anxiety and Depression Scale, the Fatigue Severity Scale, visual acuity assessment, swallowing 50 ml of water, and a clinical assessment of dysarthria.

Results: The test–retest reliability of the GNDS total score (r = 0.972) and each of its components (r varied from 0.685 to 0.987) was good. When compared with the EDSS or the Barthel, the GNDS had good validity (respectively r = 0.636 and r = –0.757). The validity was also measured for the different areas of the GNDS by comparison with corresponding indices or test. The results varied from good to excellent (r from –0.557 to 0.910). The results were only a little less reliable when a postal questionnaire was used.

Conclusion: The GNDS is a valid and reliable tool for the assessment of MS patients. The test–retest correlation varies from good to excellent. It can be used as postal questionnaire even if, as expected, the intervieweradministered method shows slightly better results.