In relapsing-remitting and secondary progressive multiple sclerosis (MS), MRI activity on monthly brain scans is a sensitive primary outcome measure in short term exploratory treatment trials. Because conventional MRI findings have a limited correlation with disability, the primary outcome in definitive trials should be clinical, although MRI is useful in providing an index of pathological progression. In trials aimed at preventing evolution from a clinically isolated syndrome to MS, MRI findings should be used in the entry criteria. The likely pathological substrates of irreversible disability are demyelination and axonal loss. Putative MR markers for these pathologies appear to relate more closely to disability than conventional MRI findings. Further technical developments should lead to improved quantitation, pathological specificity and clinical correlations.