Background: Diabetes mellitus (DM) is a major health problem with long-term microvascular and macrovascular complications responsible for the majority of its mortality and morbidity. The development and progression of diabetic complications are strongly related to the degree of glycemic control. To decrease the occurrence of these complications, instruments for self-monitoring of blood glucose (SMBG) have been developed and have become widely used among diabetic patients. In this study, we determined the relationship between SMBG control and glycosylated haemoglobin (HbA lc) levels in patients with type 2 diabetes, with and without diabetic retinopathy. Methods: Two hundred and sixty-seven type 2 diabetic patients (mean age [mean±S.D.]: 58.07±9.13 years, duration of diabetes: 8.63±6.8 years) participated in this study. Following an educational program on SMBG, glucometers and usage of oral antidiabetic agents or insulin, optic fundi were examined and HbA lc levels were measured at baseline and after 6 and 12 months. The patients were classified in three groups according to their funduscopic findings: without retinopathy ( n=140, 52.4%), background retinopathy ( n=75, 28.1%) and proliferative retinopathy ( n=52, 19.5%). Results: HbA lc levels at baseline, after 6 and 12 months were 9.09±2.69%, 7.47±1.78% and 7.12±1.4%, respectively, mean±S.D. The values decreased significantly after the education program ( P<.001 for both values compared with baseline). The prevalence of retinopathy (both background and proliferative) was 0.8% in the group of diabetics with a mean HbA lc level <6%, 7.1% in those between 6.1% and 6.9%; 9.4% in those between 7% and 7.9%; 11.8% between 8% and 8.9%; and 70.9% in those exceeding a mean HbA lc level of 9%. There was a statically significant relationship between proliferative diabetic retinopathy and body mass index (BMI; P<.001). The same relationship was observed between duration of diabetes and diabetic retinopathy ( P<.001), but not between sex and diabetic retinopathy ( P=.46). Conclusions: Implementing a program of SMBG control in type 2 diabetic patients results in lower levels of HbA lc at 6 and 12 months. In the group without diabetic retinopathy at 6- and 12-month controls, the mean HbA lc concentration is less than 7%, but in the group with diabetic (background and proliferative) retinopathy, this value could not be reduced below 7%. These results imply that SMBG would allow us to maintain better metabolic control by improving HbA lc levels and we have always kept in mind that SMBG was a part of an educational program. On the other hand, improving glycemic control prevents the onset or progression of diabetic microvascular complications, such as diabetic retinopathy, nephropathy and neuropathy. Long-term clinical studies should be performed to determine cost-effectiveness and the effects of SMBG on diabetic complications, morbidity and mortality.