OBJECTIVES: Our objective is to estimate the medical and non-medical savings associated with two smoking cessation interventions in a Minnesota health plan using a new predictive technology, and to determine the return on investment (ROI). METHODS: We developed a Markov model to simulate the smoking behaviors, health outcomes and impact of tobacco control programs in a managed care setting. We integrated the model into a software application called MediSave and used it to determine the combined impact of two interventions, which include coverage of smoking cessation pharmaceutical aids (such as buproprion or nicotine substitutes) and physician advice. Program participation and cost data were provided by the health plan. Intervention quit rates were derived from literature. Medical savings result from the higher quit rate associated with the interventions. Non-medical savings accrue to the employer because of reduction in smoking breaks as well as illness days. The ROI is calculated as the net (medical and non-medical) savings over program costs. RESULTS: In a commercially insured population of 980,292 members, the combined interventions generate positive net savings of $0.02 per member per month (PMPM) at the end of first year. They generate gross savings of $0.17 PMPM, only 33% of which are from medical savings. First year ROI is estimated to be 16% and predicted to increase significantly over the next 20 years. Eliminating non-medical savings from the analysis will result in a negative ROI in the first year but will not affect the positive net savings past year 5. CONCLUSIONS: MediSave Smoking Cessation is an effective tool in making the business case for tobacco control programs. In our study, the combined impact of pharmaceuticals and physician advice is cost-beneficial. The interventions produce substantial non-medical savings in the short term and generate significant medical savings in the long term.