Data were collected as part of a 17-month spousal assault study in a large southern city; the data source consisted of records of calls for service to the police department; the data do not include any abuse incidents not reported to police. Two tracking forms were used. The "domestic miscellaneous incident" report was a modification of an existing police form, contained minimal demographic information, and was completed on all domestic-assault cases where there was no evidence that a crime had been committed. The other form, the supplement report, was a more extensive form designed to accompany the existing police offense report, which was used when the incident involved a misdemeanor or felony; the supplement report provided data on what occurred and what action was taken. During the 17-month period, there were 18,712 domestic-violence calls for police service. Of these, 85.4 percent were judged to be noncriminal. Contrary to the popular image of serious violence perpetrated upon a spouse, the data show that most police calls involved less serious incidents that were almost as likely to involve cohabitants as married couples. This finding is significant because of the small proportion of the cohabiting population compared to the married population. Other types of relationships that generated calls to the police included, although to a lesser extent, parent-child, boyfriend-girlfriend, and sibling conflicts. Explanations for these findings focus on relationship issues and implications for service use. 53 references and 4 tables