Abstract
The present study examined the effects of random breath testing (RBT) on drink-driving crashes in New South Wales. Statistical techniques were used to distinguish "true" effects of RBT from those that were actually due to other safety interventions (such as the installation of red light cameras), or to changes in economic conditions. Particular attention was paid to any differential effects that RBT might have had on male versus female motorists, younger versus older motorists, injury versus non-injury accidents, and accidents in the Sydney metropolitan area versus accidents in other parts of NSW. Tests were also conducted to determine whether RBT's effectiveness changed since it was introduced.