OBJECTIVES: The objective of this study was to estimate the direct and indirect costs of treated and untreated insomnia in an employed population. METHODS: The Medstat MarketScan® Database was used for this study. Patients were included if they had a primary diagnosis of insomnia and/or received a new prescription for a non-benzodiazepine hypnotic medication between July 1, 1999 and June 31, 2003. Total healthcare costs, plus costs due to absenteeism, were calculated for the insomnia cohort (n=5,605), and for the propensity score matched non-insomnia cohort (total n=55,580), during 6-month pre-index and post-index periods. Change in total costs were compared using an ordinary least square model for insomnia patients who were treated versus not initially treated with a prescription hypnotic within 14 days of an insomnia diagnosis. RESULTS: Prior to matching, the insomnia cohort was slightly younger (40 vs. 42 years), more likely to be female (44% vs 31%), and had significantly more medical and psychiatric comorbidity than the non-insomnia cohort (Charlson Comorbidity Index score 0.32 vs. 0.11; P<0.01). After using propensity score matching and second stage regressions, the difference in average total expenditures in the 6-month post-index period between the cohort of insomnia patients (n=5584) and matched non-insomnia controls—the burden of insomnia—was $2738 (p<0.001). Healthcare utilization contributed to 84% of total insomnia-related costs, while absenteeism contributed 16%. Six-month costs for prescription hypnotics averaged less than $100 per patient. Both the treated and initially untreated insomnia patients experienced an increase in total costs; however, the increase for treated insomnia patients was $788 less than for the initially untreated insomnia patients. CONCLUSION: Insomnia has a significant impact on direct healthcare cost, and on costs related to absenteeism. Insomnia treatment appears to be cost-effective relative to non-treatment, or delayed treatment.