Context: Diabetes mellitus is associated with higher risk for mortality and morbidity in patients undergoing surgery. This risk is reduced if the diabetes is under control prior to the surgery. When a patient is newly detected to have diabetes during preoperative assessment a delay in controlling the blood sugar is encountered before the patient is fit for surgery. The prevalence of undetected diabetes in South India has been reported to be around 10%. Aim: We conducted the study to determine the magnitude of this problem in our surgical population. Materials and Methods: Hospital records of 529 consecutive patients who came for elective surgeries in our hospital were examined to note if they were already detected to be diabetic or were newly detected during the preanaesthetic visit. Results: 5.1% of the surgical patients were found to be diabetic. Among them 1.8% was newly detected to have diabetes. Prevalence of diagnosed and undiagnosed diabetes in patients above 40 years of age was 5.9% and 2.9% respectively. Conclusions: We conclude that the problem of undetected diabetes is not as severe as suggested by previous studies. Further studies among the general population in this geographical area is needed to know if this is due to a genuine reduction in the number of diabetics in this population.