Objective: More than one-third of patients with diabetes exhibit poor adherence with recommended drug regimens. While poor adherence is associated with excess morbidity and mortality, it is important to realize that better adherence may also affect short-term health care costs. This research quantifies the effects of improvements in medication adherence on short-term health services utilization and their associated costs. Methods: Data from Florida: A Healthy State (FAHS), a Medicaid disease management program developed jointly by Pfizer Inc. and the state of Florida targeting chronically ill Primary Care Case Management Program beneficiaries. The sample contains 5000 diabetes patients who were continuously eligible for Medicaid benefit and FAHS for at least 2 years. Medication adherence was measured with the Medication Possession Ratio (MPR) for diabetes-specific prescriptions, statins, and ACEs/ARBs. MPR was calculated separately for the first 12 months (baseline) and the following 12 months (follow-up). Costs were similarly summed over the 12 month intervals of baseline and follow-up. Estimation was performed with first-difference and conditional logistic regressions Results: Improvement in MPR between baseline and follow-up was reflected in lower inpatient, emergency room (ER), and total non-drug costs. A 10% improvement in MPR for hypoglycemic medications reduced inpatient costs by $127.97 (p=0.002), ER costs by $3.72 (p=0.003), and total non-drug costs by $116.00 (p=0.009) over 12 months. This reduction in inpatient costs was the result of a shorter average length of stay: a decrease of 0.2 (p=0.03) day per 10% increase in MPR. Finally, a 10% improvement in adherence with statins reduced the probability of being hospitalized by 4.3% (p=0.01). Conclusion: This research demonstrates the existence of significant short-term savings related to the improvement of medication adherence among persons with diabetes.