The study examined the frequency and patterns of assault within a large regional population of Victoria. The records of 860 victims of physical assault who presented to the Geelong Hospital emergency department during 1993 and 1994 were analysed retrospectively. The policy data on 1427 reported cases of physical assault, from the same catchment area and over a similar period of time, were also examined. The hospital data revealed that 65 per cent of assault victims were males aged from 15 to 34 years, that 58 per cent of presentations were within four hours of midnight, and 68 per cent were on Friday, Saturday or Sunday. Most assaults occurred on the streets, on footpaths, in open spaces or within dwellings, and the highest incidence was during the summer months. The police data showed similar patterns but from a different population of victims. It also showed an upward trend in the rates of assault in Victorian regional areas. The award-winning Geelong 'Local Industry Accord', a police and community intervention program, may have contributed to the decline in violence, particularly within the Geelong city centre, and has been suggested as a model for other community-based intervention programs. For intervention programs to be successful, the demography of assault within communities must be established and target groups identified. Hospital, police and victim surveys data do not identify accurately the population of assault victims. Improved methods of data collection and compatibility between databases already in existence may provide more accurate statistics upon which intervention programs may be based and their success evaluated.