The trend to start disease-modifying therapy early in the course of multiple sclerosis makes it important to establish whether the benign form is a real entity. In previous studies, measures of magnetization transfer (MT) ratio (MTr) have been shown to provide good estimates of the amount of tissue damage occurring in multiple sclerosis brains. Thus, with the hypothesis that if benign multiple sclerosis patients were really benign, sensitive measures of subtle tissue damage would be less pronounced in these patients than in very early relapsing-remitting (RR) multiple sclerosis patients. We carried out conventional MRI and MT imaging in 50 patients with benign multiple sclerosis [defined as having Kurtzke Expanded Disability Status Score (EDSS) <3 and disease duration >15 years] and in 50 early RR patients selected to have similar disability (EDSS <3) and short disease duration (<3 years). Data were compared with those of 32 demographically-matched normal controls. We used a fully automated procedure to measure lesional-MTr, perilesional-MTr, normal-appearing white matter (NAWM) MTr and cortical-MTr. We found that, after correction for common effects of age, lesional-MTr and perilesional-MTr of benign patients were significantly (P < 0.0001) lower than WM of normal controls, but significantly (P < 0.0001) higher than corresponding tissues of RR patients. In NAWM and cortex, MTr values of benign patients were similar to those of normal controls (P > 0.5) and significantly higher than those of the RR patients (P < 0.0001 and P < 0.01, respectively). Similar differences in MTr measures between benign and RR patients were found when patient groups were selected to have no disability (EDSS < or = 2) and, for benign multiple sclerosis, very long disease duration (>20 years) or when both groups were matched for high lesion load (T2-weighted lesion volume >10 cm3). We conclude that lesional and non-lesional MTr values can be significantly less pronounced in benign multiple sclerosis than in a cohort of RR patients at their earliest disease stages, suggesting that brain tissue damage is milder in benign multiple sclerosis than in early RR disease. This can be due to an extraordinary beneficial response to demyelination of benign patients and may represent the evidence that benign multiple sclerosis truly exists and might be differentiated from other forms of this illness.