Background/Aims: "Journey for Control of Diabetes: The IDEA Study" is an ongoing randomized trial evaluating the efficacy of group diabetes education (GE) using US Diabetes Conversation Maps(TM) and conventional individual diabetes education (IE) among adults with sub-optimally controlled Type 2 diabetes (A1c>=7%). The study is being conducted at 2 sites, Minneapolis, MN, and Albuquerque, NM. Because the 2 sites are regions with diverse demographic characteristics, the goal of this paper was to identify differences in the main outcome (A1c) between sites. Methods: In addition to A1c>=7, study inclusion criteria were patients who were diagnosed with diabetes, and were diabetes education-naive within the last one to two years. Consenting patients were randomly assigned to receive either: GE, IE or usual care (UC) as recommended by the primary care provider. All education was provided through each care system's ADA-recognized education program. Follow-up A1c levels were obtained from electronic laboratory data. Differences in A1c levels were evaluated using Generalized Linear Models (SAS 9.2). Results: There were 623 patients enrolled: 337 from MN and 286 from NM. Sample demographics were: mean age of 60.0 (MN) and 64.0 (NM); percent female, 44.8 (151/337) for MN and 54.9 (157/286) for NM; race/ethnicity, 81.1% (276/337) white, 8.1% (27/337) black, 3.6% (12/337) Hispanic, and 7.2% (24/337) other for MN, and NM 45.6% (130/286) white, 2.5% (7/286) black, 43.9% (126/286) Hispanic, and 8.1% (23/286) other. Mean baseline A1c was 8.1% for MN and 8.3% for NM. There was a significant difference in percentages of patients having an A1c in the post-randomization follow-up period (approximately 6 months), 89.0% (300/337) for MN and 67.5% (193/286) for NM. Of those with an A1c test, NM patients had a significantly greater A1c reduction with GE (-0.947 for NM vs. -0.136 for MN), IE (-1.093 for NM vs. -0.497 for MN), and UC (-0.700 for NM vs. -0.253 for MN), (p<.0001). Conclusions: Although it appears that IE and GE were more effective for NM patients compared to MN, the large number of NM patients without A1c values in the follow-up period makes any conclusions tentative. The long-term data analysis is pending and should provide a more definitive conclusion.