OBJECTIVES: This study estimated the effectiveness of rosuvastatin 20mg (R20) versus atorvastatin 40mg (A40) and 80mg (A80), and rosuvastatin 40mg (R40) versus A80 in preventing clinical events in several higher cardiovascular-risk patient populations using simulation. METHODS:   The Archimedes Model was used to simulate head-to-head clinical trials in several populations based on 10-year Framingham risk score (FRS) levels (<10, 10-20, >20) and EURO-SCORE (<5, >5) to estimate the occurrence of MACE (comprising MI, stroke, and cardiovascular death).  Simulated patients ages 45–70 with FRS >5% were drawn from the National Health and Nutrition Examination Survey.  Treatment models were validated using biomarker/outcomes data from published trials. RESULTS:   The patient numbers in each FRS and EURO-SCORE level population ranged from 9190 to 38,313.  R20 reduced MACE more than A40 or A80 and R40 more than A80 in all scenarios, with higher risk subgroups showing greater absolute benefit.  The 5-year number needed to treat (NNT) to prevent a MACE event for R40 versus A80 for EURO-SCORE <5, and >5 were 352 and 154 for 5 years and 154 and 72 for 10 years, respectively. The 5-year relative risk (RR) of MACE for R20 versus A40 was approximately 0.9, irrespective of baseline risk.  The 5-year RR of MACE for R20 versus A80 ranged from 0.92 to 0.94, and for R40 versus A80 it was 0.88 to 0.90.  RR estimates were similar at 10 and 20 years; however, NNT decreased over time. CONCLUSIONS:   The Model estimated that R20 lowers the risk of MACE more than A40 or A80, and R40 further lowers risk compared with A80.  The estimated absolute risk reduction with rosuvastatin was greater with higher baseline risk and over time.  While simulation models cannot replace controlled clinical trials, this study highlights the potential of using rigorous modeling approaches to bridge evidence gaps.