This study was designed to quantify the increased risk of road crash-related injury, which can be attributed to risk-taking behaviour. A case-control study was conducted to compare motor vehicle drivers (car and bike) who had been hospitalised for injuries following crashes with population-based controls. Cases were recruited prospectively over 12 months and controls were randomly selected from license holders (car and bike) living in the same geographical location as cases. A self-administered questionnaire was used to ascertain participants' driving behaviour, general risk-taking behaviour and selected demographic characteristics. After adjusting for demographic variables, number of years of driving and total distance driven per week, logistic regression analysis showed that a high risk acceptance was associated with an eight-fold increased risk of having a crash that resulted in serious injury (OR 7.8, 95% CI 4.2-15.8). The findings of this study support the suggestion that certain host factors increase the risk of crash-related serious injury. There would appear to be a reasonable argument for persisting with injury prevention programmes, which concentrate on host as well as environment risk factor reduction.