Background: Diabetes mellitus (DM) is an “Iceberg disease”. Worldwide there are about 150 million estimated cases of diabetes and is predicted to double by 2025. The rising prevalence of DM in developing countries is closely associated with industrialization and socio-economic development. Prevalence rates of diabetes are increasing rapidly in both urban and rural India. Hence this study has been undertaken to find at risk individuals before onset of the disease in our service area. Aim: To assess the risk of diabetes mellitus in adults ≥20 years by using Indian Diabetic Risk Score, in rural area of Pondicherry. Methods : A community based cross-sectional study conducted in a village under rural service area of Community Medicine Department. Data collected were entered and analyzed on SPSS software. Test applied were Simple proportion, Chi square test, sensitivity and specificity. Results: A total of 379 adults aged ≥ 20 years participated in the study, of which 200 (52.8%) were males and 179 (47.2%) were females. By using IDRS score 22.4% were with low risk, 48% with moderate and 29.6% were found to have high risk score, among them 1.2%, 15.4% and 34% were known to have diabetes respectively. Performance of IDRS and BMI for diagnosing diabetes mellitus showed, sensitivity and positive predictive value 56.71%, 33.92%, and 29.85%, 22.22% for IDRS and BMI respectively. Conclusion: IDRS is the better tool to detect undiagnosed diabetes when compared to BMI. It is an easy and cost effective tool which can be applied to mass for screening high risk individuals. GTT has to be done among subjects with high risk score (IDRS>60) to detect early occurrence of diabetes.