Abstract. The aim of this study was to investigate the effect of strict glycemic control on the carotid artery intima-media thickness (IMT) in type 2 diabetic patients who initially had good glycemic control (HbA1c between 5.8 and 6.4 %). The subjects were 67 patients showing deterioration of the mean HbA1c over 3 years by more than 0.2% from baseline (D group) and 33 subjects showing improvement of the mean HbA1c by more than 0.2% from baseline (A group). The clinical characteristics and annual change of IMT during the observation period were compared between the two groups in a 3-year retrospective longitudinal study. The baseline characteristics and the mean values of BMI, blood pressure, and serum lipids during the study period did not differ significantly between the two groups. However, the mean HbA1c of A group was significantly lower than that of D group (5.67 ± 0.10 vs. 6.28 ± 0.08, mean ± SE, p<0.001). The adjusted annual increase rate of IMT was significantly less in A group than in D group (-0.035 ± 0.019 vs. 0.036 ± 0.015 mm, M ± SEM, p<0.001). These results indicate that further improvement of glycemic control from a good HbA1c value can prevent an increase of IMT in type 2 diabetic patients. IT is well-known that the onset and progression of atherosclerosis is much more rapid in diabetic patients than healthy non-diabetic subjects, and one of the major causes of death in diabetes is atherosclerotic diseases. Thus, strict glycemic control is important to prevent the onset and progression of atherosclerosis. The UK Prospective Diabetes Study (UKPDS), a largescale prospective study of newly diagnosed type 2 diabetic patients, showed that every 1% decrease of HbA1c could reduce the incidence of myocardial infarction and stroke by 14% and 11%, respectively [1]. However, the UKPDS failed to show a significant reduction in the incidence of stroke and myocardial infarction in the intensive glycemic control group (median HbA1c: 7.0%) compared with the conventional glycemic control group (median HbA1c: 7.9%) over 15 years [2]. These results suggest that much stricter glycemic control and an HbA1c below 7.0% may be needed to prevent atherosclerosis in diabetic patients. The Japan Diabetes Society (JDS) had proposed an HbA1c from 5.8% to 6.4% as a marker of good control, with less than 5.7% as excellent control. However, there is no evidence as to whether the risk of atherosclerosis can be decreased by further reducing HbA1c levels from the good control range (5.8-6.4%) to the excellent range (£5.7%) in Japanese type 2 diabetic patients. Therefore, many diabetic patients with good control are maintained at such HbA1c levels, and no attempts are made to achieve excellent control. We previously reported a method of measuring carotid artery intima-media thickness (IMT) by B-mode ultrasonography and showed that the IMT of subjects with diabetes was larger than that of age-matched subjects with normal glucose tolerance [3,4]. Previous