Data on 412 victims of physical assault and 139 victims of sexual assault were drawn from the 1996 Women's Safety Survey, a nationally representative sample survey. For both physical assault and sexual assault, the survey measured attempted and threatened assaults and determined whether the last assault incident had been reported by the victim to the police. Logistic regression analysis consistently showed that assault not involving injury and assault perpetrated by a current partner were less likely than other types of assault to be reported and less likely to result in the use of victim services. In some instances, there were relatively lower reporting and service use rates for young, Australian-born, and first-time victims. Policy implications of the survey findings are discussed. 54 notes and 19 tables