OBJECTIVES: To evaluate the effect of a policy implemented in the Louisiana Medicaid program which capped reimbursement to eight prescriptions per member per month on medication persistency in patients with diabetes, hypertension, or hyperlipidemia.  METHODS: A pre-post non-equivalent comparator group design was applied using Medicaid claims data from 2001-2003 for Louisiana (LA) and Indiana (IN) to identify patients with the specified conditions and their medication persistency.  Persistency was defined as the number of days to discontinuation which was identified as a gap in treatment 30 days or longer.  To capture pre-intervention trends in medication persistency, we compared “pre-policy” cohorts in LA and IN followed for ten months prior to policy adoption  (March 3, 2002 to December 31, 2002) to policy cohorts followed ten months after policy adoption (March 3, 2003 to December 31, 2003).  All incident cohorts were identified using a six-month washout period.  We used Cox-proportional hazard models to compared discontinuation rates in LA and IN across the pre-policy and policy period cohorts. RESULTS: After adjusting for patient characteristics and comorbid conditions, no significant differences in persistency were found prior to policy implementation between LA and IN for any of the three chronic conditions. In the post-policy period, all cohorts had significantly lower persistency in LA than in IN.  Patients in LA with diabetes and hypertension were 1.38 (p= 0.03) and 2.00 (p<0.01) times more likely to discontinue their medications at day 30 of the follow-up, respectively.  The hazard ratios declined to 1.21 and 1.67 for diabetes and hypertension patients respectively after 260 days.  The hazard ratio of discontinuation for patients with hyperlipidemia in LA was constantly 1.31 (p<0.01) across the follow-up period. CONCLUSIONS: Patients with chronic conditions subject to medication caps may be vulnerable to medication discontinuation.  Policy makers need to consider carefully when implementing such policies on patients with chronic conditions.