The oligodendrocyte–myelin unit subserves saltatory conduction of the nerve impulse in the healthy central nervous system. At one time, many disease processes were thought exclusively to target the structure and function of myelin. Therefore, they were designated ‘demyelinating diseases’. But recent analyses, based mainly on pathological and imaging studies, (re)emphasize that axons are also directly involved in these disorders during both the acute and chronic phases. Another ambiguity is the extent to which these are inflammatory conditions. Here, distinctions should be made between inflammation, as a generic process, and autoimmunity in which rather a specific set of aetiological and mechanistic conditions pertain. And there are differences between disorders that are driven primarily by immune processes and those in which inflammation occurs in response to pre-existing tissue damage.With these provisos, the pathological processes of demyelination and associated axonal dysfunction often account for episodic neurological symptoms and signs referable to white matter tracts of the brain, optic nerves, or spinal cord when these occur in young people. This is the clinical context in which the possibility of ‘demyelinating disease’ is usually considered by physicians and, increasingly, the informed patient. Neurologists will, with appropriate cautions, also be prepared to diagnose demyelinating disease in older patients presenting with progressive symptoms implicating these same pathways even when there is no suggestive past history. Both in its typical and atypical forms multiple sclerosis remains by far the commonest demyelinating disease. But acute disseminated encephalomyelitis, the leucodystrophies, and central pontine myelinolysis also need to be considered in particular circumstances; and multiple sclerosis itself has a differential diagnosis in which the relapsing-remitting course is mimicked by conditions not associated with direct injury to the axon–glial unit. Since our understanding of the cause, pathogenesis and features of demyelinating disease remains incomplete, classification combines aspects of the aetiology, clinical features, pathology, and laboratory components. Whether the designation ‘multiple sclerosis’ encapsulates one or more conditions is now much debated. We anticipate that a major part of future studies in demyelinating disease will be further to resolve this question of disease heterogeneity leading to a new taxonomy based on mechanisms rather than clinical empiricism. But, for now, the variable ages of onset, unpredictable clinical course, protean clinical manifestations, and non-specific laboratory investigations continue to make demyelinating disease one of the more challenging diagnostic areas in clinical neurology.