Objectives: To examine the child welfare service response to families referred to the child welfare system in Ontario, Canada due to concerns about intimate partner violence (IPV). Methods: Bivariate analyses of a representative provincial dataset were conducted examining two types of maltreatment investigations: (i) investigations in which exposure to IPV was the only form of maltreatment; (ii) investigations in which exposure to IPV co-occurred with at least one other form of maltreatment. A stepwise logistical regression approach was used to determine statistically significant predictors of the decision to provide ongoing child welfare services. Results: Secondary data analyses of the OIS-2008 revealed that significant predictors of the decision to provide ongoing child welfare services to investigations referred by the police for exposure to IPV included whether the exposure co-occurred with another form of maltreatment, child aggression or depression and several caregiver risk factors including physical health, drug abuse, mental health issues and few social supports. Conclusions/Implications: The current approach to responding to cases of IPV is inefficient – families are referred for services multiple times but the cases are not opened for ongoing services indicating that the family’s needs are not being met. Suggestions are made for improving the child welfare service response.