Many interventions have been implemented around the world in an attempt to deter drink drivers and prevent road crashes. However, these programs may not always target high risk offenders. If high risk drivers could be identified prior to receiving convictions for drink driving, or prior to being involved in a crash, they could then be targeted for a brief road safety intervention. For example, such an opportunity would occur when a road user was in hospital receiving treatment for an alcohol-related problem. The aim of this study was to measure the association between alcohol-related medical conditions and road crashes. This was achieved through a longitudinal analysis of the hospital records of people first admitted to hospital for an alcohol-related condition over the period 1980 to 1997 in Western Australia. During this period there were 41,537 people admitted to hospital at least once for an alcohol-related medical condition. About 11.7 per cent of these people were also involved in a road crash requiring hospitalisation. However, only 23.6 per cent were admitted to hospital for a road crash after being admitted for an alcohol-related medical condition. The remainder were either involved in a crash prior to being admitted to hospital for an alcohol condition, or were treated for an alcohol condition at the same time as being involved in a road crash. The implications of these finding for early drink driving intervention will be discussed.