Background: Prevalence of Type 2 Diabetes Mellitus in Indian population is on rise and is leading to significant morbidity and mortality. The revised ADA guidelines since 2010 have suggested glycated haemoglobin (HbA1c) as a diagnostic test for Type 2 diabetes. HbA1c level is having geographic and ethnic variability independent of glycaemic status and this has not been extensively studied in many regions. Objective: To study the characteristics of HbA1C in the South Indian population and to determine the cut off value for HbA1C in diagnosing Diabetes in them. Methods: This is a hospital based observational study conducted at a Tertiary care centre in North Kerala. Patients with age ≥30yrs with Fasting Plasma Glucose (FPG) ≥ 100 or Random Plasma Glucose (RPG) ≥ 200 with symptoms of hyperglycaemia with no previous history of anti diabetic treatment were selected for the study. A preset questionnaire was used to collect the data which was later analysed using relevant statistical techniques. Results: Out of the consecutive diabetic patients attending medicine OPD 99 newly detected diabetics/prediabetics who were not under any anti diabetic medications were studied. New onset diabetes/prediabetes patients showed a mean HbA1c of 8.26(SD=2.31). There was no significant correlation of HbA1c values with age, sex and duration of diabetes in the study population. The HbA1c values also had no significant correlation with systolic blood pressure, total cholesterol values or triglyceride levels. However it was significantly related to Fasting plasma glucose, Post prandial plasma glucose and serum LDL cholesterol values. In newly detected diabetic patients (with FPG≥126 taken as gold standard) ROC analysis determined a HbA1c cut off at 6.45% (AUC=0.76, sensitivity=79%, specificity=60%). Conclusion: The study provides a reliable cutoff of glycated haemoglobin (6.45%) among South Indian population which is in accordance with the ADA recommendations.