OBJECTIVES To investigate the impact of Medicare Part D on elderly patients' compliance with statins. METHODS This is a quasi-experimental study with a pre-test and post-test for a treatment and comparison group. This study used IMS Health's LifeLink longitudinal patient level data. The subjects were elderly patients who filled prescriptions for statins in both January 2005 and January 2006. The subjects who enrolled in Medicare Part D in 2006 served as the treatment group and those who did not enroll served as the comparison group. Compliance was measured by proportion of days covered (PDC) over 11 months in both 2005 (pre-test) and 2006 (post-test). The subjects' baseline characteristics including age, gender, payment type, and patient type (statin naive vs. continuing) were compared by t-test or chi-square test. The difference-in-difference model was used to test whether Medicare Part D has an impact on patients' compliance. RESULTS A total of 473,475 elderly patients filled statin prescriptions in both January 2005 and January 2006. The mean age was 71.6 years with 55.85% female. A majority of the subjects was continuing statin users in 2005 (97.52%) and had pharmacy insurance (97.98%). A total of 201,378 (42.53%) patients enrolled in Medicare Part D in 2006. Female patients (chi-square=6101.43, p<0.0001) and patients who paid their prescriptions by cash in 2005 (chi-square=2264.82, p<0.0001) were more likely to enroll Medicare Part D in 2006. The average PDC was 80.33% and 74.40% for Medicare Part D patients and 80.04% and 71.80% for non-Medicare Part D patients in 2005 and 2006, respectively. After controlling baseline characteristics, Medicare Part D patients had 2.31% more PDC than non-Medicare Part D patients (b=2.31, t=23.69, p<0.0001). CONCLUSIONS Medicare Part D has a significant and positive impact on elderly patients' medication compliance. It is important to encourage elderly patients to enroll Medicare Part D.