BACKGROUND: Use of everolimus-eluting stents (EES) has proven to be clinically effective and safe in patients with ST-segment elevation myocardial infarction but it remains unclear whether it is cost-effective compared to bare-metal stents (BMS) in the long-term. We sought to assess the cost-effectiveness of EES versus BMS based on the 5-year results of the EXAMINATION trial, from a Spanish health service perspective. METHODS: Decision analysis of the use of EES versus BMS was based on the patient-level clinical outcome data of the EXAMINATION trial. The analysis adopted a lifelong time horizon, assuming that long-term survival was independent of the initial treatment strategy after the end of follow-up. Life-expectancy, health-state utility scores and unit costs were extracted from published literature and publicly available sources. Non-parametric bootstrapping was combined with probabilistic sensitivity analysis to co-assess the impact of patient-level variation and parameter uncertainty. The main outcomes were total costs and quality-adjusted life-years. The incremental cost-effectiveness ratio was expressed as cost per quality-adjusted life-years gained. Costs and effects were discounted at 3%. RESULTS: The model predicted an average survival time in patients receiving EES and BMS of 10.52 and 10.38 undiscounted years, respectively. Over the life-long time horizon, the EES strategy was ¿430 more costly than BMS (¿8,305 vs. ¿7,874), but went along with incremental gains of 0.10 quality-adjusted life-years. This resulted in an average incremental cost-effectiveness ratio over all simulations of ¿3,948 per quality-adjusted life-years gained and was below a willingness-to-pay threshold of ¿25,000 per quality-adjusted life-years gained in 86.9% of simulation runs. CONCLUSIONS: Despite higher total costs relative to BMS, EES appeared to be a cost-effective therapy for ST-segment elevation myocardial infarction patients due to their incremental effectiveness. Predicted incremental cost-effectiveness ratios were below generally acceptable threshold values.