Multiple sclerosis is a chronic, persistent inflammatory-demyelinating disease of the central nervous system that typically presents as an acute clinically isolated syndrome attributable to a monofocal or multifocal demyelinating lesion, which usually affects the optic nerve, spinal cord, or brainstem and cerebellum. Although the diagnosis of multiple sclerosis is still based on clinical findings, magnetic resonance imaging is now integrated in the overall diagnostic scheme of the disease because of its unique sensitivity to demonstrate the spatial and temporal dissemination of demyelinating plaques in the brain and spinal cord. Conventional magnetic resonance imaging techniques, such as T2-weighted and gadolinium-enhanced T1-weighted sequences are highly sensitive in detecting multiple sclerosis plaques and provide a quantitative assessment of inflammatory activity and lesion load. However, there is a persisting mismatch between clinical and magnetic resonance imaging efficacy of approved treatments, which underlies the fact that this technique does not suffice to explain the entire spectrum of the disease process. In recent years, great effort has been dedicated to overcoming these limitations by using non-conventional magnetic resonance-derived metrics that can selectively measure the more destructive aspects of multiple sclerosis pathology and monitor the reparative mechanisms. These metrics, which include unenhanced T1-weighted imaging, measures of central nervous system atrophy, magnetization transfer imaging, proton magnetic resonance spectroscopy, diffusion-weighted imaging, and functional magnetic resonance imaging, provide a better approximation of the pathological substrate of the multiple sclerosis plaques, have increased our understanding of the pathogenesis of the disease, and have proven useful for studying the natural history of multiple sclerosis and monitoring the effects of new treatments. Therefore, magnetic resonance imaging not only plays an essential role in the diagnosis of multiple sclerosis, but can also serve as a true biological marker of the severity of this disease.