The data included victimization incidents from the latter part of 1992 through 1994 and involved incidents in which victims reported having been attacked or threatened. Each incident involved a single victim and a single offender. The analyses excluded incidents involving robbery, rape, or burglary. The three main independent variables were the victim-offender relationship, the victim's gender, and the offender's gender. Results indicated that the victim-offender relationship affected third-party reporting but not victim reporting. The effect on third-party reporting occurred partly because third parties were unlikely to witness assaults involving people in ongoing relationships, particularly couples, and partly because third parties were reluctant to report minor assaults that involved a threat but no actual attack and no weapon. Findings indicated that when people believe that they have been the victim of a crime, their social relationship to the offender does not appear to affect whether they report the incident to the police. Tables, footnotes, and 35 references (Author abstract modified)