Cannabis is the most widely used illicit drug in the world. There is strong evidence from performance studies that delta9-THC has significant effects on the cognitive and psychomotor tasks associated with driving. New, easy and sensitive methods to detect impaired drivers are needed. Therefore, it is necessary the use of alternative biological samples which may be accurate, precise and with trustful interpretation results. This article presents an LC-MS methodology for detecting delta9-tetrahydrocannabinol (delta9-THC) in oral fluid samples. The mean recovery was 79%, coefficients of variations were between 2.9% and 6.9% and the limits of detection (LOD) and quantitation (LOQ) were 1.0 ng/ml and 2.0 ng/ml, respectively. The method is sensitive, accurate and reproducible and may be utilized in ongoing controlled cannabinoid administration studies and in roadside studies and thus, important for the fields of forensic toxicology.