Kurtzke's EDSS remains the most widely-used measure for clinical evaluation of MS patients. However, several studies have demonstrated the limited reliability of this tool. We introduce a computerized instrument, MS-CANE (Multiple Sclerosis Computer-Aided Neurological Examination), for clinical evaluation and follow up of patients with multiple sclerosis (MS) and to compare its reliability to that of conventional Expanded Disability Status Scale (EDSS) assessment. We developed a computerized interactive instrument, based on the following principles: structured gathering of neurological findings, reduction of compound notions to their basic components, use of precise definitions, priority setting and automated calculations of EDSS and functional systems scores. An expert panel examined the consistency of MS-CANE with Kurtzke's specifications. To determine the effect of MS-CANE on the reliability of EDSS assessment, 56 MS patients underwent paired conventional EDSS and MS-CANE-based evaluations. The inter-observer agreement in both methods was determined and compared using the kappa statistic. The expert panel judged the tool to be compatible with the basic concepts of Kurtzke's EDSS. The use of MS-CANE increased the reliability of EDSS assessment: Kappa statistic was found to be 0.42 (i.e. moderate agreement) for conventional EDSS assessment versus 0.69 (i.e. substantial agreement) for MS-CANE (P=0.002). We conclude that the use of this tool may contribute towards a standardized and reliable assessment of EDSS. Within clinical trials, this could increase the power to detect effects, thus reducing trial duration and the cohort size required. Multiple Sclerosis (2000) 6 355 - 361