The Domestic Violence Research Project was a pilot study designed to examine the dynamics of domestic violence within two of the ten police districts that comprise the city of Boston. The objectives were to collect data on domestic violence in greater detail than previously possible, conduct various analyses on this information, and determine how the findings could best be used to improve the police, prosecutorial, and social service responses to domestic violence. Data for 1993 are a stratified random sample of reported domestic violence incidents occurring throughout the year. The sample represents approximately 27 percent of the domestic violence incidents reported in 1993 for the two districts studied, B3 and D4. The 1994 data include all reported incidents occurring in the two districts during the period May to July. After the incident selection process was completed, data were collected from police incident reports, follow-up investigation reports, criminal history reports, and court dockets. Variables include arrest offenses, time of incident, location of incident, witnesses (including children), nature and extent of injuries, drug and alcohol use, history of similar incidents, whether there were restraining orders in effect, and basic demographic information on victims and offenders. Criminal history information was coded into five distinct categories: (1) violent offenses, (2) nonviolent offenses, (3) domestic violence offenses, (4) drug/alcohol offenses, and (5) firearms offenses.