OBJECTIVES: To develop a budget impact model estimating potential costs of adding new drugs to the formulary using medical claims and clinical trial data using a case study of eszopiclone, an insomnia medication awaiting approval. METHODS: A 3-step process estimated the costs of treating insomnia using eszopiclone and other insomnia agents: 1) Insomnia drug efficacy was estimated from published literature on marketed drugs and from clinical trial reports of eszopiclone. Prior to summarizing trial data, adjustments were made for the different characteristics of each clinical trial (i.e., size, location, patient characteristics); 2) A multiple regression model was then applied to the MarketScan claims data (from 85,832 insomnia patients) and to the clinical trial efficacy data in order to estimate a relationship between medical expenditures and treatment efficacy. Patient-reported total sleep time was the primary efficacy measure, and five other sleep variables (patient-reported latency and awakenings, PSG latency, sleep efficiency, and wake time after sleep onset were included in the model; 3) Eszopiclone expenditures (premium pricing assumed) were estimated by multiplying eszopiclone efficacy estimates by the regression-based estimates of expenditures. The regression model additionally accounts for the impact of other demographics and clinical characteristics that influence expenditures. A budget impact model that could be used for any given health plan was then developed. RESULTS: For a health plan with 100,000 members, the model estimated that use of eszopiclone decreased annual insomnia treatment costs by $498,204 using the default efficacy endpoint, total sleep time. Results were consistent regardless of which efficacy endpoint was used. CONCLUSIONS: Using this modeling approach, use of eszopiclone is estimated to substantially decrease annual treatment costs for insomnia patients. The modeling approach used represents an alternative and potentially effective way of estimating the impact of medications pending marketing approval.