Objectives: To compare predicted coronary heart disease (CHD) events and lipid goal attainment under the Framingham and a managed care database risk equation in a managed care patient cohort with established CHD. Methods: Independent outcomes models were developed from the Framingham-based risk equation (FR model) and from a risk equation using a large Integrated Research Database (IRD model). Prior CHD patients =50 years (distribution based on NHANES data) with combined sub-optimal LDL-C (=100 mg/dL), HDL-C (=40 mg/dL, males; =50 mg/dL, females), and TG levels (=150 mg/dL) at baseline were modeled to predict CHD events and lipid goals attained over five years. IRD model covariates included age, gender, follow-up time, combined lipid goal achievement, and Deyo-Charlson co-morbidity index. CHD events included acute coronary syndrome, sudden coronary artery disease death, and cardiac catheterization (PCI). Patient therapy was based on a formulary with all major branded and generic lipid drugs. Results: Using a hypothetical 1,000,000 member plan, 35,059 patients aged =50 years were identified with prior CHD. The percent of patients achieving LDL-C and TG goals over five years was higher in the IRD model versus FR model (68.9% vs. 62.7% and 69.9% vs. 49.2%, respectively), while HDL-C goal achievement percent was comparable (53.4% vs. 54.5%). The number of patients experiencing a CHD event over five years was higher in the IRD versus FR model (17,194 vs. 6,387), reflecting the inclusion of actual practice CHD intervention (i.e., PCI) in unstable angina (14,430 vs. 1,461). Conclusions: The established FR model appears to conservatively underestimate both lipid response and theoretical CHD events, nor does it account for CHD clinical intervention. The IRD model may better reflect the real world lipid response and CHD events, also accounting for actual event indicators of CHD intervention, being developed from a treatment population.