Background: There is considerable variation in the ability of physicians to bring diabetic patients under glycemic control. The purpose of this study was to examine medication prescribing patterns relative to glucose control of the clinically most successful doctors in the management of diabetes patients. Methods: The data investigated came from data collected from the electronic medical record (EMR) of HealthPartners Medical Group (HPMG) between 3/1/2005 and 2/28/2006. The study looked at 80 primary care physicians treating diabetes patients in 20 clinics around the twin cities area. Physicians were ranked on the number of diabetes patients at the combined goals of having an A1c <7%, a low density lipoprotein (LDL) <100 mg/dl, blood pressure <130 mm, a nonsmoker and taking a daily aspirin. The top 25% (n=20) of these physicians were classified as ‘optimal physicians’ and compared with the remaining 75% of physicians (n=60). Medication adjustments were examined at each primary care encounter relative to the last known A1c level. Results: There was no difference in the optimal physician distribution between the physicians having a specialty of family medicine (47%) relative to those with a specialty of internal medicine (P=0.1299). The numbers of diabetic patients treated (mean 100) was also not a factor associated with optimal physician status (P=0.9361). Optimal physicians managed more patients to their glucose goals (23.1% vs. 18.9%; P=0.0290). When we looked at the average A1c level where physicians continued to titrate the dose of oral glucose control medications for patients above goal, there was a significantly (P=0.0120) lower mean threshold for titration for the optimal verses the other physicians (7.8% vs. 8.4%). Conclusions: While all physicians are focused on managing diabetes patients who are above their glucose goal, the optimal physicians are willing to continue making medication adjustments to a point closer to the glucose goal than other physicians. This does not appear to be related to experience gained through having more diabetes patients in their patient panel, but may indicate that optimal physicians are more comfortable or perhaps knowledgeable about continually titrating medications when patients are near the glucose goal. Perhaps if other physicians could be given further training and support tools such as what may be embedded within an EMR, they could also achieve similar results.