Objective: To understand the interaction between health systems and individual factors in determining the probability of a cesarean delivery in India. Methods: In a retrospective study, data from the 2007-2008 District Level Household and Facility Survey was used to determine the risk of cesarean delivery in six states (Punjab, Delhi, Maharashtra, Andhra Pradesh, Kerala, and Tamil Nadu). Multilevel modeling was used to account for district and community effects. Results: After controlling for key risk factors, the analysis showed that cesareans were more likely at private than public institutions (P < 0.001). In terms of demand, higher education levels rather than wealth seemed to increase the likelihood of a cesarean delivery. District-level effects were significant in almost all states (P < 0.001), demonstrating the need to control for health system factors. Conclusion: Supply factors might contribute more to the rise in cesarean delivery than does demand. Further research is needed to understand whether the quest for increased institutional deliveries in a country with high maternal mortality might be compromised by pressures for overmedicalization. (C) 2014 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.