OBJECTIVES: In India, prescriptions are kept by the patient and not with the pharmacist; and antibiotics may be obtained with or without a prescription. Therefore, determining antibiotic use is problematic, since there are no prescription records. The objective of this study was to develop a surrogate measure for antibiotic use in the community in India. METHODS: Surveillance of antibiotic use was conducted in different residential areas that fall under five municipal wards of New Delhi, India during December'03-December'04. Thirty retail pharmacies were enrolled for the study and consumption of antibiotic use (bulk purchase data) was collected by recording the quantities of all antibiotics purchased by these pharmacies every month over one year. ‘Exit interviews' were conducted with patients leaving the enrolled pharmacies. 15-25 exit interviews were conducted for each pharmacy per month over one year. Consumption of antibiotics was defined as total number of Defined Daily Dose (DDD)/1000 population and DDD/1000 patients. The denominator for bulk data was population of the ward and for exit interviews, number of patients who visited the pharmacy. RESULTS: Bulk purchase data and exit interview data showed a similar pattern of antibiotic use. Consumption of antibiotics belonging to the fluoroquinilone group was much higher than others. Penicillins, cephalosporins and macrolides showed almost equal usage in the community. Exit interviews revealed that 22% of persons who visited the pharmacies bought antibiotics. DDD/1000 patients for fluroquinolone, penicillins, cephalosporins and macrolides were 34493, 15029, 13270 and 11800 respectively. Newer drugs were used more than older drugs. CONCLUSION: Both drug use methods are good indicators of antibiotic use in the community. In future, we will use the exit interviews, since this method is less dependent on commitment of the pharmacists as well as being a means to collect additional information on patient perceptions and prescriber behavior.