The aim of this review was to determine the spectrum and severity of effects of unintentional antipsychotic poisoning in children. A computerised literature search of MEDLINE (1966 to February 2005) and EMBASE (1980 to February 2005) was undertaken. The Internet was searched using URL: www.google.com. The proceedings of the North American Congress of Clinical Toxicology (NACCT) and the European Association of Poisons Centres and Clinical Toxicologists (EAPCCT) were hand searched. All cases of unintentional antipsychotic (all classes) poisoning in children aged 0-6 years were included. The data extracted included the age, weight, antipsychotic, dose, clinical effects, treatment and outcomes. The toxic dose was estimated as the lowest dose causing objective adverse effects.Sixty-eight reports were identified. Few contained all of the required information. Most of the case series included multiple antipsychotics with limited information on individual drugs or all ages with limited paediatric information. For most antipsychotics the ingestion of one tablet caused symptoms that were sometimes severe and usually lasted from 1 to 3 days. Extrapyramidal symptoms (EPS) were often delayed for up to 12-24 hours. Chlorpromazine caused CNS depression, hypotension and miosis; EPS and cardiac effects were rare, and the toxic dose was estimated to be 15 mg/kg. Haloperidol caused drowsiness (rarely coma) and over one-half of patients had neuromuscular effects (mainly EPS), with a toxic dose estimated at 0.15 mg/kg. Thioridazine caused CNS depression and potentially cardiac effects, with a toxic dose of 1.4 mg/kg. Atypical antipsychotics caused significant CNS depression (except risperidone); EPS were less common. Toxic doses were clozapine 2.5 mg/kg, olanzapine 0.5 mg/kg and aripiprazole 3 mg/kg. EPS responded to anticholinergic drug treatment. In summary, unintentional antipsychotic ingestion in children can cause severe effects that last 1-3 days, often with one tablet. Children potentially ingesting a toxic dose or who are symptomatic should be considered for assessment in hospital. Most cases resolve with good supportive care. Toxic doses are only estimates that are based on limited data and should be used with caution until prospective studies are undertaken.