The study found that the majority of substantiated family-violence offenders attacked their spouses and had not abused their children (61 percent), followed by child abusers who had not committed spouse abuse (27 percent) and 12 percent who had committed offenses against both their spouses and their children. These three groups of family-violence offenders differed in the types of abuse they perpetrated (neglect of children, emotional abuse, physical abuse, and sexual abuse); their experiences of being a spouse abuse victim; and sociodemographic characteristics. Twelve percent of all spouse abusers had committed multiple abuse incidents; and 10 percent of all child abusers had committed multiple incidents of child abuse. The authors urge caution in interpreting these findings due to the study's methodological limitation (many incidents of spouse abuse and child abuse go undetected and unreported). Despite these limitations, however, these findings provide descriptive information that may help in informing practice and policy among those who develop and deliver family-violence prevention and therapeutic services. Study data were obtained from the Army Central Registry, which is a confidential electronic information system maintained by the U.S. Army Medical Command. The Registry obtains its data from the Army's Family Advocacy Program, which is primarily responsible for family-violence prevention, identification, evaluation, treatment, and followup on Army installations. 4 tables and 35 references