India accounts for 1.7 million child deaths, a quarter of global child mortality. The current literature has succeeded in establishing an association between domestic violence and child mortality, but has yet to present evidence of a causal relationship. In this paper we use an instrumental variable approach to analyse the causal impact of domestic violence against the mother on child mortality in the Indian context. Domestic violence is instrument ed with the real price o f gold at the time of marriage of the mother. Results lend evidence to a bias in OLS estimates and s how a significant positive relationship between domestic violence and mortality. A one standard deviation increase in domestic violence translates to a 6 percentage point increase in both neonatal and infant mortality.