OBJECTIVES: This study assessed the predictors of compliance and persistence to statin therapy among Texas Medicaid patients. METHODS: A retrospective cohort-analysis using pharmacy and medical claims of new statin users between September 1, 1999-August 31, 2001 was done. Compliance was assessed using medication possession ratio (MPR), which is a ratio of the sum of the days' supply (except on the last prescription refill date) to the total number of days between the first and last prescription refill date. Persistence was defined as the time to therapy discontinuation or the failure to refill a prescription within 60 days of exhausting the last supply. The predictors of compliance and persistence that were assessed included gender, age, ethnicity, presence of disease conditions, and total number of prescriptions other than statins. Logistic regression and Cox regression models were used to assess the predictors. RESULTS: Of the total (N=7440) patients, 65.2% were females and the mean age was 49.7 years (S.D.=9.4 years). White, non-Hispanic (42.7%), Hispanic (32.7%), and Black, non-Hispanic (22.5%) formed the majority of the ethnic categories. The mean MPR was 0.7 (S.D.=0.2) and 56.7% of the patients had an MPR of less than 0.8. At the end of 310 days, only 50% of the patients were still persistent with their therapy. The cumulative probability of being persistent to therapy at the end of the 2-year period was 0.41. The predictors of improved compliance and persistence to statin therapy were male gender, being non-Hispanic white, and absence of disease conditions such as CHD, diabetes, and hypertension. CONCLUSIONS: Adherence to statin therapy is suboptimal. There exists a gender gap as well as ethnic disparity with regards to adherence. Steps need to be taken to promote adherence in order to reduce the risk and the long-term costs associated with CHD.