OBJECTIVES: To quantify the impact on medical costs of having an integrated medical and pharmacy benefit in the Diabetes Mellitus (DM) population. METHODS: Patients who had at least two diabetes ICD-9 codes or two diabetes medications refills between January 1, 2006 and December 31, 2006 were included. Patients with a diabetes diagnosis or diabetes medications within one year before the index date were excluded to ensure a treatment-naïve DM population. Only patients continuously enrolled from January 1, 2006 to December 31, 2011 were included in this retrospective cohort study. The Gemod model with Gamma distribution was utilized to adjust for the baseline disease comorbidity, age, gender, and account type differences. A total of 2090 patients met the eligibility criteria: 1087 in the integrated group (Members using an integrated medical and pharmacy benefit) and 1003 in the non-integrated group (Members using medical and pharmacy benefits from separate providers). RESULTS: The average unadjusted annual mean medical costs were $7,299 in the non-integrated group vs. $5,561 in the integrated group (p<0.01). From year 1 to year 5, the growth of adjusted mean medical costs was 53% in the non-integrated group compared to 47% in the integrated group. The adjusted mean medical costs were higher in the non-integrated group (N=1003) than in the integrated group (N=1087) in each of the five years (p<0.01). While treatment naïve DM patients tended to have relatively low inpatient and emergency room costs, in each year, the integrated groups’ mean inpatient and emergency room costs were on average $568 lower when compared to the non-integrated group. This observed difference was not statistically significant. CONCLUSIONS: The DM patients in the integrated group had lower medical costs and lower inpatient and emergency room costs than their non-integrated counterparts. These results suggest a value to integrated health benefits; further research is required to elucidate the drivers of these observed savings.