This retrospective study examines the effect of a medication assistance program (MAP) on HbA1c levels in an uninsured, low-income, type 2 diabetes population. It also examines the degree to which improvement in HbA1c level varied with adherence to medication regimens among those patients using the MAP. The MAP was found to have a mean effect of -0.60% on HbA1c levels. However, MAP users differed in how strictly they adhered to medication regimens, as measured by number of refill opportunities taken. The MAP's effect on HbA1c varied monotonically with adherence level, with greater adherence leading to greater HbA1c improvement. Never refilling the prescription (complete nonadherence) led to no change in HbA1c, while complete adherence led to an estimated -0.88% improvement in HbA1c. Further study is needed to investigate factors related to non-adherence within medication assistance programs and the effect of such programs on other patient outcomes.