OBJECTIVES: To investigate the association between insomnia, health care utilization and cost in a representative population. METHODS: The America Insomnia Survey (AIS) is an epidemiology study of insomnia prevalence and burden of illness in the US. The Brief Insomnia Questionnaire (BIQ), a fully structured diagnostic interview, was administered to a representative sample of subscribers from 14 geographically dispersed US health plans in the HealthCore Integrated Research Database. The aim was to investigate associations between insomnia and relevant outcomes, (daytime impairment, accidents, injuries, lost productivity, health care consumption). Respondents were medical and pharmacy benefit eligible and had minimum12 months continuous enrollment prior to the survey date. Survey and administrative claims data were merged. Data were analyzed using SAS Genmod with log link and gamma distribution for costs adjusting for Deyo-Charlson Index score. Respondents were classified as ‘any insomnia (AI)’ and ‘no insomnia (NI)’ with ‘any insomnia’ defined as meeting criteria for any of 3 nosologic systems (DSM, ICD, Research Diagnostic Criteria [RDC]). RESULTS: Of 10,094 respondents, insomnia prevalence is 23.6%.  Of insomnia subtypes alone and in combination, waking in the night is most prevalent (61.5%), followed by waking too early (52.1%), difficulty falling asleep (38.0%) and non-restorative sleep (6.2%). Compared to NI, AI are more often female (58.1% vs. 49.6%; p<0.0001) and younger (45.1±16.6 vs. 46.6±17.6; p=0.001). Preliminary claims analysis suggests within 12 months prior to the survey, AI had significantly higher all-cause medical costs than NI respondents ($4,830 vs. $3,714, respectively; p<0.0001). All-cause pharmacy costs were significantly higher for AI than NI  ($1,186 vs. $959, respectively; p<0.0001). CONCLUSIONS: The prevalence of insomnia is higher than previously thought with sleep maintenance as the most important symptom.  Insomnia accounts for significant health care costs in the US.