Results indicate that there was a disproportionately low arrest, charge, and prosecution rate in domestic violence cases compared to other violent cases: 15 percent of male domestic violence was not prosecuted compared to 7 percent of other male violence cases. It has been argued that this differential treatment stems from witness withdrawal or unreliability. However, these problems are found in other types of violence cases, and the decision to prosecute is not dependent on witness withdrawal. Similarly, a comparison of the two types of cases indicates that domestic violence cases are rarely dropped because they are trivial, and relatively trivial cases of both types often proceed to prosecution. Victim precipitation of the offense also does not appear to produce a satisfactory explanation for differential treatment: only one domestic case was not pursued for this reason, while two nondomestic cases were dropped because of it. A possible explanation may be found in police attitudes that prosecution in domestic cases will not be effective, that the police role is one of crisis intervention, and that policing is primarily concerned with the preservation of public order. It is argued that differential treatment of domestic violence stems from these views, a perception that family matters are more a civil than a criminal matter, and that the police mandate for order maintenance contributes to prioritizing of public over private violence cases. 3 tables, 8 footnotes, and 52 references.