Background:Diagnosis of child mania has been contentious.Objective: To investigate natural history and prospective validation of the existence and long-episode duration of mania in children.Design: Four-year prospective longitudinal study of 86 subjects with intake episode mania who were all assessed at 6, 12, 18, 24, 36, and 48 months.The phenotype was defined as DSM-IV bipolar I disorder (manic or mixed) with at least 1 cardinal symptom (elation and/or grandiosity) to ensure differentiation from attentiondeficit/hyperactivity disorder.Parent and child informants were separately interviewed, by highly experienced research nurses, using the Washington University in St Louis Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS).A Children's Global Assessment Scale score of 60 or less was needed to establish definite impairment.Treatment was by subjects' community practitioners.Setting: Research unit in a university medical school.Participants: Subjects were obtained from psychiatric and pediatric sites by consecutive new case ascertainment, and their baseline age was 10.8 ± 2.7 years.Onset of the baseline episode was 7.4±3.5 years.(Data are given as mean±SD.)Main Outcome Measures: Episode duration, weeks ill, recovery/relapse rates, and outcome predictors.Results: Prospective episode duration of manic diagnoses, using onset of mania as baseline date, was 79.2±66.7 consecutive weeks.Any bipolar disorder diagnosis occurred during 67.1%±28.5% of total weeks, during the 209.4±3.3 weeks of follow-up.Subjects spent 56.9%±28.8% of total weeks with mania or hypomania (unipolar or mixed), and 38.7%±28.8% of these were with mania.Major or minor depression and dysthymia (unipolar or mixed) occurred during 47.1%±30.4% of total weeks.Polarity switches occurred 1.1±0.7 times per year.Low maternal warmth predicted faster relapse after recovery from mania ( 2 =13.6,P=.0002), and psychosis predicted more weeks ill with mania or hypomania (F 1,80 =12.2, P=.0008).Pubertal status and sex were not predictive.(Data are given as mean±SD.)Conclusions: These findings validate the existence, longepisode duration, and chronicity of child mania.Differences from the natural history of adult bipolar disorder are discussed.