Although homicide rates have remained fairly stable in Australia for the last 20 years, police figures show large increases in most forms of nonfatal violent crime. Further, victimization surveys show no evidence of even a slight increase in violent crime, and some data show a decrease. Given these discrepancies, there is reason to consider the possibility that the increase in violent crime reflected in the police figures is not the result of changes in the "real" level of violence in the community, but rather the result of a range of factors that can be categorized as "police productivity." Police productivity can be taken to include a range of factors likely to enhance the performance of the police in detecting and recording offenses. This paper suggests that homicide levels are a good indicator of levels of violence generally, and that since homicide data are reliable and show no increase in the rate of homicide, it follows that trends in nonlethal violence are similar. There is sufficient evidence to conclude that the popular view that violence in Australia has increased dramatically and consistently in recent years is unfounded. On the basis of existing information, it is possible to identify some factors that are "violence facilitators," such as cultural attitudes toward violence, alcohol consumption patterns, and the availability of firearms. To reduce the violence rate, public policy should focus on undermining the acceptability of violence at all levels of society as well as identifying the risk factors that are associated with violence. 4 figures and 14 references