OBJECTIVES To construct a flexible Markov model allowing cost-effectiveness analysis of different statin therapies over different time horizons in ACS patients in Germany. METHODS A Markov model was designed to compare outcomes of intensive atorvastatin 80 mg versus standard simvastatin 20–40 mg treatment. Event rates for the first 5 years for MI, stroke, unstable angina, heart failure, revascularization and cardiovascular death were derived from a post-hoc analysis of a subset of patients with recent ACS in the IDEAL study; extrapolation of event rates beyond the trial was based on risk-prediction equations from the Framingham study. Event and all-cause mortality were based on sources in literature. To estimate QALYs, utilities were derived from the literature. Cost inputs were based on the German statutory health insurance perspective; patients' co-payments were deducted. Costs of care for cardiovascular events were based on DRG costs for acute treatment, and on literature and expert panel results for subsequent costs. Univariate and probabilistic sensitivity analyses were performed. All costs and benefits were discounted by 5% annually, and costs were reported in 2008 Euros. RESULTS The base case used a one-year cycle length, a lifetime time horizon, and a five-year treatment duration with high-dose atorvastatin, after which all patients received simvastatin for life. For the base case, the incremental cost per life-year gained is €13,993, and the incremental cost per QALY gained is €14,168. At a threshold of €30,000 per QALY gained, sensitivity analysis predicted an 85% probability that atorvastatin 80 mg would be considered cost-effective compared with simvastatin 20–40 mg. CONCLUSIONS The results can only approximate real-life clinical practice; however, the carefully constructed model and selection of input parameters, combined with data based directly on a post-hoc analysis of a clinical trial, make this a useful contribution to the debate regarding the incremental benefit of intensive statin therapy.