Indonesia’s rapidly growing mega-urban regions of Jabotabek (Jakarta Extended Metropolitan Region), Bandung and Surabaya each have a large socially and economically heterogeneous population. As a result, substantial spatial and socio-economic inequalities in child mortality risk exist. A means for policy makers to address these differentials is to increase the utilization of trained birth assistance, especially to the urban poor. Multilevel multinomial logistic regression reveals that use of a skilled birth attendant is highly associated with predisposing factors (ie, mother’s education) whilst highly skilled assistance is more related to enabling factors (ie, household expenditure quintile and proximity to hospitals). Recent
decentralization of government powers, however, has caused numerous challenges to policy makers to ensure more equitable access to health services and reduce inequalities in child
mortality risk.