OBJECTIVES: Obesity (defined as body mass index (BMI) > 30) is a major contributor to increased morbidity, mortality, and health care expenditures.  We used data from the Medical Expenditure Panel Survey (MEPS) to construct a lifetime cost and outcomes model explicitly accounting for morbid obesity.  Then, we estimated the potential economic value of weight reduction in terms of cost savings plus the value of improvements in life expectancy and quality of life. METHODS: We constructed a Markov model with a lifetime horizon using death and BMI health states:  18.5-24.9, 25-29.9, 30-34.9, 35-35.9, 40-44.9, 45-49.9, 50<.  MEPS Panels 6-10 (2001-2006) provide 91,000 observations to estimate transition probabilities. We estimated discounted (3% per annum) lifetime costs and quality-adjusted life years (QALYs) using regressions with gender, age, and BMI as predictors.  Utilities were mapped from SF-12 to EQ-5D. RESULTS: In the base case,  for a 45 year-old female with BMI 45-49.9, a 10%, 20%, and 30% BMI reduction is associated with discounted lifetime economic value gain of $20,000, $35,000, and $51,000, respectively.   Transition probability estimates show that patients tend to stay in the same BMI category from year to year.   Average annual cost increases significantly (p<0.01) with BMI (+$362 per 5 BMI unit increase), age (+$118 for each year of age) and gender (+$547 for females).  For utility levels, BMI (-0.0246 per 5 BMI unit increase), age (-0.0036 for each year of age), and gender (-0.0355 for females) were all significant (p<0.01). CONCLUSIONS: MEPS panels provide repeated measures of BMI allowing a projection of the natural history of obesity from a single data source. Obesity is a significant driver of costs and QALYs:  we show that a reduction in BMI of obese patients could be associated with lower costs and better quality-adjusted life expectancy.  MEPS provides only self-reported BMI data, and this potential bias deserves further exploration.