Background: Among ASEAN countries, Indonesia is the country with high maternal mortality ratio (MMR) and unable to reach target of reduction of MMR. In order to reduce the MMR target, the government of Indonesia needs evidence to evaluate and design the maternal health program. Objective: The study aims to answers specific issues of who, when, where and why maternal death occurred in Indonesia in order to understand the effective policy and health program decisions. Methods: The 2010 Indonesia Population Census identified pregnancy-related deaths occurring in the household from 1 January 2009 until the date of census (May 2010). The follow-up study revisited almost half of households reporting pregnancy-related deaths to be accounted as samples (4167 of 8464). Basic information related to cause of death were collected by trained data collector using verbal autopsy approach. The information was converted to cause of death defined by medical doctor using WHO ICD-10 rules. The underlying cause of death was later analysed. Result: The highest risk of maternal death was adolescents who were pregnant under 15 years old. The maternal death mostly occurred at postpartum period (56%), 57 percent occurred at hospitals and 31.3 percent at home. Oedema, proteinuria and hypertensive disorder in pregnancy were at 27 percent, whereas complication during labour and delivery problems were accounted for 26 percent. The pattern of maternal causes of death varies between regions. Conclusion: Maternal health program has not been considered as general plan of intervention. It is imperative to consider considered by pattern of characteristics and cause of maternal death and region for effective interventions.