Though studies have focused on the expansion of coverage of health care utilization after India’s health care reform in 2004–2005 under the National Rural Health Mission, the effect of limited risk protection of the programme, measured by non-availability of services at institutions, has remained under-studied. Using bivariate probit models on birth history data of 6568 children in two districts in West Bengal, an eastern state of India, this paper finds that in post-reform era, health care utilization in the form of institutional delivery has a tendency to enhance the risk of infant mortality, primarily owing to lack of timely services received at the institutions. © 2016 Informa UK Limited, trading as Taylor & Francis Group.