Magnetic resonance imaging (MRI) has revolutionized the diagnosis and management of patients with multiple sclerosis (MS). Conventional MRI metrics are employed as primary endpoints in proof-of-concept clinical trials evaluating new drugs for MS and as secondary endpoints in definitive phase III trials. Metrics derived from non-conventional MRI techniques are now emerging and hold significant promise since they appear to be more correlated with the most disabling features of MS. However, none of these has been approved for use as a surrogate endpoint for accumulation of physical disability, which is the most important clinical endpoint of this disease. Taking into account the large numbers of patients needed, the extensive exposure to placebo, and the relatively long duration required for phase III clinical trials to show a meaningful effect on progression of disability, the need for a valid, reliable, and objective paraclinical marker of disease evolution cannot be overemphasized. This paper reviews the most up-to-date data regarding MRI techniques, their relationship with central nervous system pathology, as well as with clinical endpoints, and proposes future insights into the use of MRI metrics as surrogate endpoints in clinical trials of MS.