Sibutramine and orlistat are effective in obesity (a major independent risk factor for several diseases). High obesity prevalence, questionable long-term benefits, and drug costs create a dilemma for insurers on these drug-coverage decisions. OBJECTIVE: To determine whether sibutramine or orlistat is cost-beneficial in obesity treatment. METHODS: Randomized placebo-controlled trials evaluating sibutramine or orlistat for weight-loss were collected from MEDLINE and IPA searches. Inclusion criteria were patients 18 years and older with BMI 27-44kg/m2. Exclusion criterion was comorbidities. After assessing quality, using common endpoints at 6 months, 1, and 2 years, 5 trials on orlistat (120mgTID) and 6 on sibutramine (10mgQD) were selected. Pooled estimates of weighted mean difference (between placebo and drug) in proportion of patients losing 10 % or more of body weight (the primary outcome) were derived. Lifetime health and economic benefits of sustained 10% weight loss for adults aged 35-64 years with mild (BMI 27.5kg/m2), moderate (BMI 32.5kg/m2) and severe (BMI 37.5kg/m2) obesity from a recently published cost-of-obesity model (Am J Public Health, 1999,89:1536) of the relationship between BMI, risks, and cost of five obesity-related diseases (hypertension, hypercholesterolemia, diabetes, coronary heart disease, stroke) were used. Using direct costs, CBA was performed with weighted average savings (transformed to year 2001 US$) by age, BMI, and proportion losing 10 % or more of body weight. Sensitivity analysis by varying the proportion of patients sustaining 10% weight loss (+/-20%) and discount rate was done. RESULTS: Benefit-cost ratios of orlistat and sibutramine were less than one for the base-case analysis at 6 months,1 year and 2 years. Best/worst-case benefit-cost ratio (via sensitivity analysis) of orlistat remained less than one, while that of sibutramine was 1.8 /1.2 at 1 year. CONCLUSIONS: Unless the proportion of patients sustaining 10% weight-loss increases, neither orlistat nor sibutramine would prove cost-beneficial in formulary coverage decisions.