OBJECTIVES: Many drugs show long-term rather than short-term benefits. Moreover, most clinical studies were too short to give evidence of cost savings. Therefore, we developed a simulation model to estimate the costs and benefits of BMI lowering drugs within a time frame of 10 years. We applied this simulation, a Markov-model, on data from 4429 patients (20 studies) treated with the weight-management agent sibutramine or placebo. Both groups also received diet and exercise advice. METHODS: As major savings of lipid and BMI lowering drugs will be induced by reductions in hypertension, CHD and diabetes (H-CHD-D) incidence rates we restricted our model to these diseases. We chose a time frame of 10 years. In order to estimate changes in the incidence rates of H-CHD-D we used risk-equations data derived from large epidemiological studies. Our Markov model contains the following disease states: no disease; only hypertension; diabetes without CHD; CHD, and death. We estimated the transition probabilities by using the respective risk equation. Finally, each state was linked to annual costs. Using this simulation, we could estimate the individual costs for hypertension, CHD, and diabetes for each of the patients within the time frame of 10 years. The mean difference of these costs between the sibutramine and placebo group were considered as savings induced by the sibutramine therapy. RESULTS: The 10 year incidence rates (sibutramine vs. placebo) were 22.3% vs. 33.8% (hypertension), 4.5% vs. 6.1% (CHD), and 19.4% vs. 21.3% (diabetes). The mean costs induced by H-CHD-D were 8042 DM for sibutramine and 9671 DM for placebo, a reduction of 1629 DM. CONCLUSION: Considering annual sibutramine costs of 912 DM a net benefit of 717 DM over 10 years per patient results. This is a conservative estimate as only three diseases related to obesity were included.