The data came from the National Crime Victimization Survey during 1987-1992. The research also examined models predicting various police responses to a victim's report, including police response times, police actions at the scene, and the probability of an arrest. Results revealed that a combination of contextual and demographic characteristics affected the probability that incidents would be reported. Black victims, victims who were injured, and victims who reported that the offender had not victimized them before were more likely than others to report their victims. The same factors were also significant predictors of arrest. Police were more likely to make arrests in incidents in which victims had sustained injuries, when the offender did not have a history of violence, and when the incident involved black offenders victimizing black victims. Findings indicate that the victims who have experienced repeated victimizations and are therefore at greatest risk of sustaining injury are least likely to report the victimization and that repeat offenders, who are more likely to cause an injury, are less likely than others to be arrested. Tables, notes, appended table, and 40 references (Author abstract modified)