Cognitive impairment in multiple sclerosis is difficult to study because of the heterogeneity and variability of this disease. The gold standard for measurement of cognitive function in multiple sclerosis is a full battery of neurocognitive tests, which is time consuming and expensive. Some cognitive tests like the PASAT, a measure of working verbal memory and processing speed, have been proposed for screening and follow-up of cognitive function in clinical trials. We studied whether we could measure cognitive function in multiple sclerosis over the Internet. For this we used the Cognitive Stability Index (CSI)™, developed for persons with known or suspected primary central nervous system illness. The CSI was compared with formal neurocognitive testing (NPsych) and the PASAT in a cross-sectional study of 40 consecutive multiple sclerosis patients with subjective cognitive complaints. NPsych revealed that only 18 of the 40 patients (46%) were cognitively impaired. Although both the CSI and the PASAT were equalivalent in their specificity (86%), the CSI was significantly more sensitive than the PASAT (83% versus 28%). We conclude that the CSI, because of its availability over the Internet, has great potential as a tool for screening and follow up of cognitive function in multiple sclerosis. Multiple Sclerosis 2007; 13: 1011—1019. http://msj.sagepub.com