To determine the association between some demographic features ( age, sex, level of education) and the rate of diabetes mellitus control, we performed a cross sectional study in 1065 patients who were under the care of Kerman diabetes center in 1992. We evaluated the data of 525 patients who had more than four visits annually. Sixty nine per cent of them were female and 31℅ male. The mean age for women was 51 and for men 52.2 years. Majority of them (45.9℅) were illiterate or low educated, 32.6℅ had primary or junior high school education, 15.6℅ high school and 5.9℅ college education. Seventy four per cent of patients had regular follow up visits within 3 months, of whom 72℅ were females. Patients were divided into three categories according to the level of diabetes control as: well, moderate based on CBS, 2 hr pp and 4 PM BS. Although the level of education was higher in men, the rate of diabetes control and regular follow up visits were better among women. There was a direct relationship between age and well controlled diabetes in both sexes, but the level of education had no association with the rate of diabetes control.