The cause of multiple sclerosis remains unknown after more than a century of study. Unconfirmed work has once more indicated that a viral infection may be important in the aetiology of the disease, and there is considerable evidence for an important genetic influence on disease susceptibility. The clinical course is as variable as that of any disease in medicine. Studies using serial magnetic resonance imaging have helped to define the disease course and response to experimental therapies. Although the predominant pathological characteristic is myelin loss with preservation of axons, some studies recall classic descriptions that irreversible axonal destruction may occur, perhaps even in the early stages of the illness. There are now several, partially effective therapies for relapsing forms of multiple sclerosis and here I review progress in determining the timing and course of the illness and the steps that need to be taken to identify more effective treatments for this disease.