Victim, incident, abuser, and defendant characteristics were examined for 663 restraining order cases. Each defendant was tracked for 2 years to determine further abuse which was defined as a criminal complaint anywhere in Massachusetts for abuse against the same victim. Because victims were not interviewed to determine new abuse, findings of reabuse were admittedly conservative. Although abusers were brought to civil court by their victims, the majority physically assaulted their victims and another one-third threatened to kill or otherwise harm victims, their children, or a relative. Police officers were involved in over 10 percent of cases and arrested the abuser for the incident that formed the basis of the restraining order request. The age of abusers ranged from 17 to 70 years, just over half were either married or divorced from their victims, and 80 percent had prior criminal histories. All women came to court to obtain an initial temporary order that was valid for 5 days. The overwhelming majority (75 percent) returned for the contested hearing, and all women who requested were given a longer-term order valid up to 1 year. No statistically significant differences were found between victims who maintained their restraining orders and those who asked that restraining orders be dropped early. Almost half of the abusers (48.8 percent) reabused their victims within 2 years of the 1990 restraining order; 34 percent were arrested for abuse. The only statistically significant predictors of new abuse, as measured by a new arrest, were age, prior criminal history, and court-ordered contact provisions. Younger abusers reabused more than older abusers, male batterers brought to court for civil restraining orders were active criminals, and restraining orders were clearly not sufficient to protect women and children from continued abuse. 20 references, 8 notes, 3 tables, and 8 figures