Magnetic resonance imaging (MRI) is increasingly being used as a measure of pathological disease activity in monitoring the efficacy of potential new treatments for multiple sclerosis. A major advantage of MRI over clinical monitoring is that it detects a large amount of subclinical disease activity. The two main approaches to MRI are detecting active lesions and measuring total lesion load. In relapsing-remitting and secondary-progressive multiple sclerosis, gadolinium enhancement increases the number of detectable active lesions and also probably correlates with pathological activity. Total brain-lesion load can be measured from computerized images, either by manual outlining of lesions or by more fully automated lesion-segmentation strategies. An important limitation of MRI monitoring is that conventional brain MRI abnormalities often show little or no relationship with clinical disability. One explanation for this may be pathological heterogeneity of lesions that all look the same on a conventional image. Other MR techniques are needed that specifically identify the pathological features most likely to result in disability, namely demyelination and axonal loss. Magnetization transfer imaging and proton MR spectroscopy are two techniques that show promise in this regard.