By the late 1970's, the classic response to domestic violence was widely seen by feminists, battered women's groups, and police administrators as an abject failure. A 1981 Minneapolis study changed the focus from common knowledge that domestic violence is a problem to a lively debate over which treatment option is most effective. Historically, the response to domestic violence has exhibited several characteristics: few domestic violence cases have been formally processed by the criminal justice system; police officers have avoided intervention in domestic violence cases; and there has been a strong bias against the use of arrest. Researchers confirm that only about 10 percent of domestic assaults are reported to the police. Further, those who do report do not necessarily reflect the widespread nature of domestic violence. Research also shows that most police officers do not like to intervene in domestic violence cases. Several reasons explain this reluctance: organizational impediments, lack of appropriate training, cynicism (belief that domestic violence calls are not "real" police work), and personal safety worries. Research into when arrests occur indicates weak correlations between actual arrest rates and victim injuries. Actual arrest data suggest that situation-specific factors predominate in the decision to arrest, including police/offender interactions, police perceptions of victim conduct, and police organizational issues. Many studies conclude that arrest does not necessarily fail to work in domestic violence cases, but arrest for the purpose of deterrence may not always work. The impact of an expanded police role in domestic violence cases is discussed, and directions for further research into domestic assault prevention are noted. 27 references