In the past decade much research concerning the impact of cannabis use on road safety has been conducted. More specifically, studies on effects of cannabis smoking on driving performance, as well as epidemiological studies and cannabis-detection techniques have been published. As a result, several countries have adopted driving under the influence of drugs (DUID) legislations, with varying approaches worldwide. A wide variety of bodily fluids have been utilized to determine the presence of cannabis. Urine and blood are the most widely used matrices for DUID legislations. However, more and more publications focus on the usability of oral fluid testing for this purpose. Each matrix provides different information about time and extent of use and likelihood of impairment. This review will focus on the practical aspects of implying a DUID legislation. The pros and cons of the different biological matrices used for Δ(9)-tetrahydrocannabinol screening and quantification will be discussed. In addition, a literature overview concerning (roadside) cannabinoid detection, as well as laboratory confirmation techniques is given. Finally, we will discuss important issues influencing interpretation of these data, such as oral fluid collection, choice of cut-offs, stability and proficiency testing.