OBJECTIVE To evaluate the prevalence of NIDDM and IGT in the urban and rural areas in southern India. RESEARCH DESIGN AND METHODS Two populations of the same ethnic background, but different socioeconomic background were chosen for this study. Nine-hundred urban people and 1038 rural subjects were studied. Fasting and 2-h post-glucose capillary blood samples after a 75 g oral glucose load (WHO criteria) were obtained in these randomly selected adults (≥ 20 yr of age). RESULTS Using the WHO criteria, the prevalence of NIDDM, adjusted to the age of the respective general population, was 8.2% in the urban and 2.4% in the rural populations. The prevalence was 8.4 and 7.9%, respectively, in urban men and women, and 2.6 and 1.6% in rural men and women. The age-adjusted prevalence of IGT was 8.7 and 7.8% in the urban and rural areas, respectively. The prevalence of IGT was 8.8% in urban men and 8.3% in women; the corresponding values for rural men and women were 8.7 and 6.4%. The prevalence of NIDDM increased with age, markedly so in the urban people. The urban-rural difference was significant for NIDDM (X2 = 29.4, P < 0.001) but not for IGT. In the urban population, 65% of the NIDDM patients were known cases, whereas in the rural area, the known cases accounted for only 24%. Bivariate analysis showed an association of BMI, STR, and WHR with prevalence of NIDDM plus IGT. In the multiple logistic regression analysis, age, BMI, STR, and WHR were associated significantly with glucose intolerance in the urban population, whereas only age was significant in the rural population. The best predictors of NIDDM were age, BMI, WHR, and urbanization. CONCLUSIONS The study showed a high prevalence of NIDDM in the urban southern Indian population. The prevalence of NIDDM in the same ethnic group in rural areas was significantly lower. The prevalence of IGT was similar in both populations. Upper body adiposity was a significant predictor of NIDDM in this population with low rates of obesity.