The police agency has about 120 full-time sworn police officers and had a turnover of 4.5 percent during the 5-year study period. The research focused on a random sample consisting of every 10th police report from the 92,000 police reports from 1982 through 1987. The 180 cases of domestic assault and 335 other assault cases were analyzed. Results revealed that police officers are less likely to arrest in domestic violence cases than in other assaults. In addition, injuries were about equally likely to be present in cases of arrest for domestic and nondomestic assault, although victims were more likely to experience minor injuries in domestic assaults. Thus, while victims of domestic violence cases are more frequently victims of minor violence, the level of injuries by itself has relatively little to do with arrest in either domestic or nondomestic assaults. It also appeared that police officers were more willing to consider domestic assaults as real assaults only when weapons were used. Further research is recommended. Tables, notes, and 36 references