The findings indicate that a child's perception of an interspousal violent incident as being a personal threat to himself/herself and being powerless to control the outcome or effects of the incident increased the likelihood that the child would suffer significant traumatic symptoms and behavioral problems. This finding held after controlling for the effects of demographic factors and the characteristics of the violence. An incident that also involved abuse of the child ("co-victimization") also increased the likelihood of the child having clinically significant levels of traumatic symptoms compared to children who witnessed a domestic-violence incident but were not themselves victimized. Being female and White increased the odds of specific trauma symptoms (emotional disorders) and behavior problems. Increasing age reduced the odds of some trauma symptoms. Future research should continue the effort to identify the mechanisms through which differences in exposure to and experiences of violence differentially affect children. The 687 children were all participants in a community-based program that targets children who have witnessed violence. The children selected for the study had all been exposed to domestic violence. Data were collected by mental health specialists during family visits, typically 1-2 weeks after the index violent event. Eleven survey instruments were used to measure three groups of predictors: demographics of the child and the child's family, characteristics of the violent event, and the child's perceptions of the violent event. Traumatic symptoms were identified by the Trauma Symptom Checklist for Children, and problem behaviors were identified by the Revised Behavior Problem Checklist. 4 tables and 58 references