OBJECTIVE: We sought to evaluate the quality of care in known diabetic patients from the middle- and high-income group populace of Delhi. RESEARCH DESIGN AND METHODS: A cross-sectional survey was conducted using a probability proportionate to size (systematic), two-stage cluster design. Thirty areas were selected for a house-to-house survey to recruit a minimum of 25 subjects (known diabetes >/= 1 year; aged 35-65 years) per area. Data were collected by interview, by blood sampling, and from medical records. RESULTS: A total of 819 subjects (of 1,153 eligible) were enrolled from 20,666 houses. In total, 13.0% (95% CI 9.6-17.3) of the patients had an HbA(1c) (A1C) estimation and 16.2% (13.5-19.4) had a dilated eye examination in the last year, 32.1% (27.5-36.6) had serum cholesterol estimation in the last year, and 17.5% (14.2-21.5) were taking aspirin. An estimated 42.0% (37.7-46.2) had an A1C value >8%, 40.6% (36.5-44.7) had an LDL cholesterol level >130 mg/dl, and 63.2% (59.6-66.6) had blood pressure levels >140/90 mmHg. CONCLUSIONS: A wide gap exists between practice recommendations and delivery of diabetes care in Delhi.