OBJECTIVES: The purpose of this study was to evaluate the relationship between various diabetes treatment regimens and 1) attainment of hemoglobin A1C (HbA1c) treatment goal, and 2) treatment costs. METHODS: Pharmacy, medical, and laboratory claims data from a large health plan in the Southeastern US from 1/1/2003 - 7/31/2006 were used. The index date was defined as the date of the last available HbA1c value, an indicator of glycemic control during the previous 90 days. The study period was defined as the 120-day period that preceded the index date. Goal was defined as an HbA1c <7.0. Treatment regimens were categorized based on the number of diabetes-related medications: 1 oral, 2 orals, =3 orals, insulin, insulin plus 1 oral, insulin plus 2 orals, and insulin plus =3 orals. Logistic and OLS regression models were used to identify predictors of goal attainment and total treatment costs, respectively, after adjusting for potential confounding variables. RESULTS: A total of 18,905 patients were included in the study, of which 50.9% were at goal. The highest proportion of patients at goal was in the 1 oral group (63.1%), while insulin plus 1 oral had the lowest (22.0%). After adjusting for potential confounders, patients treated with regimens that included insulin had a higher HbA1c value than those only treated with oral medications. Those treated with insulin were less likely to be at goal compared to those treated with one oral (OR=0.18, p<0.0001). Similar results were found in other insulin treatment groups. Predictors of increased total cost included male sex, higher co-morbidity scores, and treatment with multiple oral medications compared to one oral medication. Increased compliance was associated with goal attainment (p<0.0001) and lower total cost (p<0.004). CONCLUSION: Only 50% of patients attained HbA1c goal. Increased compliance was associated with goal attainment and lower costs.