The primary purpose of this study is to determine whether child protective services investigators apply a recognizable set of criteria to substantiate batterers and victims of battering for exposing their children to domestic violence. Although domestic violence occurred in 35% of the 1,248 substantiated incidents of child maltreatment, only 31 (7.1%) couples were investigated for exposing a child to domestic violence or failing to protect a child from domestic violence. All of the batterers investigated and in the caregiver role when their children were exposed to domestic violence were substantiated. The unsubstantiated victims of battering tended to use more protective behaviors (M=3.82) than the substantiated victims (M=2.00); yet, at the case level, using more than one protective behavior did not seem to be a criterion used to substantiate the victims. Instead, it appears that investigators were discriminating between those protective behaviors by the victims that ended contact between the batterers and the children--for a substantial amount of time--and those that did not in both the substantiation and removal decision. Key issues related to applying criteria in incidents involving domestic violence are discussed along with recommendations to further refine and document them. (Contains 4 tables.)