This study was carried out along with the 1990 6th National Tuberculosis Prevalence Survey in order to estimate the prevalence rate of diabetes mellitus among a representative adult Korean population. Fasting blood glucose was measured by diastix (Ames) using glucometer II for seventy seven percent of the population (30 year old or above) residing in 190 enumeration districts randomly sampled from 146,944 general ED. The prevalence of diabetes mellitus of the population was estimated by projecting the ratio of ${geq}200mg/dl;PP_2$ of fasting blood sugar level below 120 mg/dl by sex to the study population. Fasting blood glucose and 2 hr. postprandial blood glucose were measured on about 3000 subsampled individuals, and diabetes mellitus was defined by the WHO criteria-FBG${geq}120mg/dl;or;PP_2{geq}200mg/dl$ when FBG is below 120mg/dl. The results obtained are as follows : 1. Estimated prevalence(age adjusted) of diabetes mellitus was 4.6% for male and 8.1% for female. The age adjusted mean FBG was $93.9{pm}26.2mg/dl$ for male and $102.9{pm}31.5mg/dl$ for female. 2. The prevalence increased as age advanced with peak in $60{sim}69$ years old age group for both sexes. 3. The mean FBG and estimated prevalence rate of diabetes mellitus varied considerably among the populations of fifteen cities and provinces ; it vaired from $87.0{pm}17.7;to;104.6{pm}34.5mg/dl$ and 1.2% to 8.9% for males, and from $93.3{pm}25.3;to;116.7{pm}38.6mg/dl$ and 3.4% to 20.8% for females. 4. The mean FBG and prevalence rates were higher among the rural population than those residing in metropolitan areas. 5. The proportion of the people previously treated among the diabetics was estimated to be a little less than one fifth. 6. Factors strongly associated with FBG were age, sex, family history of D. M., BMI, area and educational level among eleven variables.