Magnetic resonance imaging (MRI) plays a central role in the management of patients with multiple sclerosis (MS). T2-weighted/FLAIR lesions have been included in the diagnostic criteria since 2001, and the importance of the technology has been expanded in each successive revision of the McDonald criteria. While the typical focal hyperintense lesions seen on T2 and FLAIR sequences in several areas of the central nervous system are key features for MS diagnosis, they can also be used to monitor disease activity, particularly in asymptomatic patients, and to evaluate therapeutic responses. The development of new lesions, particularly in medullary and infratentorial locations, is a strong predictor of long-term disability and risk of evolution to a secondary-progressive phase. Yet, changes in T2 lesion volume are poor predictors of subsequent disease evolution in many cases, a situation often referred to as the “clinicoradiological paradox”. Nevertheless, advanced MRI techniques allow quantification of several pathological processes in vivo and offer insights into MS pathophysiology beyond white matter lesions. By investigating what is happening beneath the visible surface of MS pathology, these techniques not only help to unravel the clinicoradiological paradox, but also provide early measures of functional and structural tissue abnormalities before the advent of irreversible neurodegeneration.