OBJECTIVES To describe utilization patterns of lipid lowering drugs (LLD) and to examine the relationship between LLD use and all-cause hospitalizations, emergency room visits and mortality among Medicare patients with diabetes. METHODS This is a longitudinal retrospective cohort study. The study sample consisted of diabetic Medicare Part D beneficiaries from six states, including Alabama, California, Florida, Mississippi, New York and Ohio. Beneficiaries was identified as diabetic using a combination of ICD9 code for diabetes (‘250.xx') and NDC codes for insulin or oral hypoglycemic agents. Utilization of LLDs in Medicare Part D records during the first 6 months of 2006 was evaluated and the measure was dichotomized as not using and using LLD. The primary outcomes of interest were all-cause hospitalization, all-cause emergency room (ER) visits and all-cause mortality. Multivariable regression analysis was performed to assess the association between LLD utilization and outcomes. RESULTS The study sample constituted of 1,888,682 Part D enrollees with diabetes. Fifty-one percent of them were users of LLD. Of the users, 59% were female, 84% were over 65 years of age and 67% were white. In unadjusted analysis, patients who used LLD had lower all-cause hospitalizations (OR: 0.761, 95% CI 0.754-0.767), lower ER visits (OR: 0.909, 95% CI 0.903-0.914) and lower all-cause mortality (OR: 0.567 95% CI 0.559-0.574). When adjusted for age, gender, race and comorbidities in multivariable analysis, LLD utilization remained significantly associated with lower risks for all-cause hospitalizations (OR: 0.862, 95% CI 0.854-0.869), lower ER visits (OR: 0.801 , 95% CI 0.796-0.806) and lower all-cause mortality (OR: 0.646, 95% CI 0.637-0.655). CONCLUSIONS In conclusion, approximately 50% of the patients with diabetes received LLDs in the study period and utilization of LLDs are shown to be associated with improved outcomes in patients with diabetes after controlling for patient demographics and comorbidities.