Driving following illicit drug consumption ('drug-driving') is a potential road safety risk. Roadside drug testing (RDT) is conducted across Australia with the dual aims of prosecuting drivers with drugs in their system and deterring drug-driving. We examined trends over time in self-reported past six-month drug-driving among sentinel samples of regular drug users and assessed the impact of experiences of RDT on drug-driving among these participants.Data from 1913 people who inject drugs (PWID) and 3140 regular psychostimulant users (RPU) who were first-time participants in a series of repeat cross-sectional sentinel studies conducted in Australian capital cities from 2007 to 2013 and reported driving in the past six months were analysed. Trends over time were assessed using the χ2 test for trend. Multivariable logistic regressions assessed the relationship between experiences of RDT and recent drug-driving, adjusting for survey year, jurisdiction of residence and socio-demographic and drug use characteristics.The percentage of participants reporting recent (past six months) drug-driving decreased significantly over time among both samples (PWID: 83%  vs. 74% , p<0.001; RPU: 72% vs. 56%, p<0.001), but drug-driving remained prevalent. Lifetime experience of RDT increased significantly over time (PWID: 6%  vs. 32% , p<0.001; RPU: 2% vs. 11%, p<0.001). There were no significant associations between experiencing RDT and drug-driving among either PWID or RPU.Although there is some evidence that drug-driving among key risk groups of regular drug users is declining in Australia, possibly reflecting a general deterrent effect of RDT, experiencing RDT appears to have no specific deterrent effect on drug-driving. Further intervention, with a particular focus on changing attitudes towards drug-driving, may be needed to further reduce this practice among these groups.