Cognitive impairment may be a common even at the onset of multiple sclerosis (MS). In this case-control study, we tried to find out the probable relationship between homocysteine levels and cerebral atrophy or cognitive impairment in patients with multiple sclerosis. One hundred fifty six patients who had MS according to McDonald diagnostic criteria were included in this study. Patients’ age, gender, and educational level, MS duration and clinical type, disability, cognitive function state based on minimental state examination (MMSE), presence of hyperhomocysteinemia, and brain atrophy were evaluated. There was no statistically significant relationship between hyperhomocysteinemia and cognitive status. Total homocysteine levels had a significant correlation with MMSE score only in those patients with elementary level of education. Also total homocysteine levels and overall cerebral atrophy did not indicate significant relationship according to those independent variables mentioned above except in the patients with EDSS less than 6. When intercaudate ratio > 0.10 was applied as a criterion for cerebral atrophy, we found that hyperhomocysteinemia related significantly to intercaudate ratio > 0.10 in females, aged between 21 and 30 years, MS duration ≤ 5 years, primary progressive MS and relapsing-remitting MS clinical types, EDSS ≤ 3 and elementary level of education. We suggest applying MMSE only for the first step of cognitive function survey. In the next steps, much more exact test must be used (e.g. MSNQ). Also we can not suggest measuring plasma homocysteine level as criterion for monitoring the cognitive function in patients with MS.