Our goal is to assess the effect of caregiver vulnerabilities, singly and in combination, on the substantiation of child abuse (physical, sexual) and neglect, while controlling for relevant background variables. We test the moderator role of adult partner violence in qualifying the relationship between caregiver vulnerabilities and maltreatment substantiation. Secondary analyses of the 1998 Canadian Incidence Study of Reported Child Maltreatment (CIS) are used to predict child protective service investigation substantiation versus non-substantiation from a range of caregiver vulnerability factors. Involvement in partner violence was examined as a moderator in the relation between caregiver vulnerabilities and maltreatment substantiation. The CIS is an epidemiological survey of first-reported cases to child protective services, using a random sample of child welfare agencies across Canada. Child welfare workers completed a research form on the child, primary caregiver, family, perpetrator, severity and type of maltreatment, as well as services and court outcomes. All maltreatment classifications were assigned according to the Canadian legal definition of child abuse and neglect. Hierarchical logistic regression analyses were used, with stepped entry of: (1) demographic factors, socioeconomic disadvantage, and caregiver's own history of maltreatment; (2) caregiver vulnerability factors; (3) involvement in partner violence; (4) the interaction between caregiver vulnerability and partner violence. Caregiver substance abuse was found to be the single most potent kind of caregiver vulnerability in predicting maltreatment substantiation. When the total number of vulnerabilities was used as the predictor, prediction across all types of maltreatment increased, especially for substantiated neglect. Analyses also showed that the presence of partner violence in the home exacerbated the effect of caregiver vulnerability on substantiation. The total number of caregiver vulnerabilities was the best predictor of the substantiation of child abuse and neglect. This relationship was moderated by the existence of partner violence: high caregiver vulnerability and high partner violence increased the likelihood of substantiation versus non-substantiation. These results suggest that caregiver issues should be considered in tandem with partner relationships. Among child welfare cases, caregiver vulnerability and partner violence are critical targets for child maltreatment prevention and early child protective services intervention.