Despite increased attention to the problem of tardive dyskinesia (TD), many questions remain unresolved. There is a consensus that neuroleptics play a substantial role in its development, but other variables must also contribute. Prevalence surveys have helped to define the scope of the problem and suggest risk factors for further study. Their usefulness has been limited by methodological problems including the difficulty of estimating false-negative (masked) dyskinesia and false-positive (movements caused by other neuromedical conditions) rates. A recent large scale survey reported an overall rate of abnormal involuntary movements of 23.4% among neuroleptic-treated psychiatric patients; the range was from 12.3% among outpatients at a Veterans Administration hospital to 37.4% among state hospital inpatients. Rates of covert dyskinesia, obtained by withdrawing medication from 70 TD negative cases, ranged from 17% for the Veterans Administration to 67% at the state hospital (overall rate, 34%). Very few clear false-positive cases were found. The incidence of TD, based on a large prospective study of young adult patients, is 19% after 4 years of cumulative neuroleptic exposure. Higher incidence rates have been found in prospective studies of older patient samples. Age remains the risk factor most consistently associated with TD development; it may also relate to increased persistence. Female sex among older populations, diagnosis of affective disorder, and evidence of neuroleptic-induced pseudoparkinsonism also relate to increased risk. Further work is needed to elucidate the role of dosage and length of neuroleptic treatment, as well as other potential contributory factors.