OBJECTIVES: Medication non-compliance has been a growing concern for healthcare management and can result in progression of cardiovascular disease (CV) and an increase in economic burden. To develop models to predict the risk for future medication non-compliance among patients with hypertension and hyperlipidemia. METHODS: Two predictive models were constructed using pharmacy and medical claims from 2380 patients newly treated with anti-hypertensive medications and 3387 patients newly treated with statins in a managed care setting. The outcomes of interest were the future medication compliance rates for both disease states over a one-year follow-up period. The potential predictors of compliance included patient characteristics such as age, gender, type of insurance plan, Chronic Disease Score (CDS), presence of select comorbidities, copayments, total medication burden, hospital encounters, outpatient physician encounters, and initial compliance (0-3 months of therapy immediately before the follow-up period). Linear regression models were applied to construct the models. Each population was randomly split by a 2 to 1 ratio to facilitate spilt-sample validation of the models. RESULTS: Based on the hypertension model, age, gender, total co-payments, total medication burden, and initial compliance showed significant relationship with compliance (R2=0.45). Based on the hyperlipidemic model, age, gender, presence of a second CV condition (e.g. angina), outpatient physician encounters, co-payments at drug initiation, total medication burden, and initial compliance demonstrated significant relationship with compliance (R2=0.43). In both populations, initial compliance was the strongest predictor of sustained compliance. CONCLUSION: These models can serve as a useful tool to guide providers in promoting medication compliance. Both models suggest that assisting the patient to establish compliant behavior within the first three months of a new treatment regimen can significantly influence sustained medication adherence with CV medications.