The study found that children were present in the home when almost half of the domestic-violence crimes occurred; and 81 percent of these children were directly exposed to the violence. Children under the age of six were at greater risk of such exposure. The domestic-violence households with children were more likely to consist of non-White adults with low incomes and to be headed by a single female compared to households in general. Based on the data collected by police officers, seven profiles of domestic violence events were developed. Highs and lows on the following variables distinguished the types: level of injury to the victim, means of assault, relationship status between victim and perpetrator, substance use during the event, mutual assault, and arrest of the perpetrator. The study found that the three types of domestic violence that had disproportionately high rates of child sensory exposure were the most dangerous and unstable types. Children had a greater likelihood of being exposed to violent events that involved injuries and weapon use, to events that involved mutual assault between intimate partners, and to events that involved substance abuse (victim or perpetrator) and arrest of the perpetrator. In collecting data on domestic-violence events to which police responded, police officers used the Domestic Violence Event Protocol-Child Enhanced. Data were collected on 1,560 substantiated domestic violence events in 2001. The resulting database was maintained by a large county police department in the Northeast. 1 table and 39 references