The usefulness of non-contrast-enhanced, standardized magnetic resonance imaging for the longterm follow-up of MS patients was evaluated in a retrospective study in 36 patients with clinically definite MS. All had remitting-relapsing diseases courses. Sixteen patients remained clinically stable during follow-up. Mean duration of follow-up was 22 months (SD: 11). A mean number of 3 MRI examinations was performed in each of the patients (SD: 1). Subclinical evolution was detected in 56% of the stable patients, indicating that clinical data alone are insufficient to assess disease activity. The relapsing patients showed significantly more and larger changes on MRI than stable patients (P < 0.001), indicating that MRI is well suited as a follow-up parameter in conjunction with clinical data. The time courses of these quantitative changes and of the qualitative changes of putative MS lesions on MRI are discussed. It is concluded that MRI is a good indicator of global disease activity in multiple sclerosis patients, which makes MRI very useful for the evaluation of therapeutic trials.