Clinical, histopathological, and laboratory evidence suggest that dancing eye syndrome/opsoclonus-mycoclonus syndrome is an immune-mediated and probably autoimmune disorder. There is a clear, but not invariable, association with neuroblastoma or ganglioneuroma in children, and with a miscellany of neoplasms in adults. Direct support, i.e. the identification of antibodies against cell-surface antigens in neurons, is lacking but the nature of other evidence is such that the case for use of immunomodulatory therapeutic strategies is compelling. Because at least 85% of children affected by the condition are left with permanent cognitive deficits, notwithstanding a favourable initial neurological response to, say, steroid or adrenocorticotrophic hormone, therapeutic stakes are high. There is an urgent need for controlled therapeutic trials which, because of the comparative rarity of the condition in children, will require international collaboration to expedite studies.