OBJECTIVES: To examine the relationship between sustained glycemic control and health care costs among patients with diabetes with an initial hemoglobin A1c (HbA1c)>9%. METHODS: We conducted a retrospective analysis of administrative data from patients with diabetes and initial poor HbA1c control enrolled in a large health plan in Hawai‘i (n=1304).  We used propensity scores to identify a comparable cohort based on age, gender, type of coverage, diabetes duration, number of medications, location of residence, comorbidity conditions, and morbidity level.  We examined the relationship between reduced HbA1c values and costs in the same year as well as the impact of achieving sustained HbA1c control (at < 7%) for three years on changes in health care costs using generalized linear models. RESULTS: In cross-sectional comparisons, the average annual direct medical costs for patients with a HbA1c less than 7% was $14,821 compared to $12,108 for the matched sample of patients with HbA1c greater than or equal to 7%, for a difference of $2,713 [$285 - $5,140].  In contrast, when we examined the change in cost from 2006 to 2009 for patients who had sustained levels of HbA1c at <7% for all 3 years, we found that total cost care for patients with sustained control decreased by $2,207 compared to a $3,006 increase for patients without sustained control, for a difference of -$5,214, 95%CI[-$10,163 - -$264]. CONCLUSIONS: Our study suggests that while reducing HbA1c levels to target goals may not immediately result in cost reductions, sustained HbA1c control is likely to reduce costs in a three-year time frame.