OBJECTIVES: This analysis estimates the impact of switching from rosuvastatin to atorvastatin on cardiovascular (CV) events. METHODS: Using the Archimedes model, a discrete-event simulation, we evaluated a simulated population of 50,038 patients aged 45–70 years with dyslipidemia and adherent to rosuvastatin. The virtual population was created based on matched biomarker profiles of patients in the National Health and Nutrition Examination Survey (NHANES). Statin treatment models were based on data from published studies, including STELLAR, JUPITER, CARDS, ASCOT, and TNT. Patients were started on a rosuvastatin dose based on their LDL-C and ATP-III goal.  Treatment prevalence was calibrated to match distributions of statin use observed in U.S. pharmacy claims data. In the experimental arm, patients switched to twice the mg atorvastatin dose as their rosuvastatin dose at their first follow-up visit (no switching occurred in the control arm). Patients were monitored for 5 years, during which time they received regular visits, with the opportunity to increase their dosage if they were above their LDL-C goal. Rate of first occurrence of MI, stroke, or CV death (MACE) for each arm was estimated. RESULTS: After 5 years, when all patients switched to atorvastatin, 4.8% fewer patients reached goal, and mean LDL-C increased by 7.1 mg/dL. The 5-year relative risk of MACE was 1.117 (95% CI 1.098–1.135), and the number needed to harm (NNH) to incur one additional MACE was 262, favoring rosuvastatin. For diabetics, the relative risk of MACE was 1.128 (95% CI 1.096–1.159), and the NNH to incur one additional MACE was 195, indicating greater risk for diabetics. The results were insensitive to assumptions of adherence rates and LDL-C goal values. CONCLUSIONS: In this simulation, switching from rosuvastatin to atorvastatin led to fewer patients attaining LDL-C goal, higher mean LDL-C values, and greater risk of MACE over 5 years.