Saliva or "oral fluid" has been presented as an alternative matrix in the establishment of drug exposure. The noninvasive collection of a saliva sample, which is relatively easy to perform and can be achieved under close supervision, is one of the most important benefits in a driving under the influence situation. Moreover, the presence of delta9-tetrahydrocannabinol (THC) in oral fluid is a better indication of recent use than when the drug is detected in urine, so there is a higher probability that the subject is experiencing pharmacological effects at the time of sampling. At 3 check points organized by the Swiss police in Bern, 61 drivers were tested for the presence of drugs of abuse using the Drugwipe 5 device. In parallel, oral fluid was collected with the Intercept DOA Oral Specimen Collection device and tested by gas chromatography-mass spectrometry (GC-MS) after methylation of THC (limit of quantitation 1 ng/mL). The Drugwipe device identified 1 exposed driver, but with GC-MS, 18 drivers tested positive. THC concentrations in the Intercept buffer ranged from 2.1 to 205.1 ng/mL. These concentrations represent about 1/2 to 1/3 the authentic THC concentrations in oral fluid because of the dilution by the blue liquid of the device. Two main limitations of oral fluid were 1. the amount of matrix collected is smaller when compared to urine and 2. the levels of drugs in urine are higher than in oral fluid. A current limitation of the use of this specimen for roadside testing is the absence of a suitable immunoassay that detects the parent compound in sufficiently low concentrations.