Oral fluid (OF) testing offers simple, non-invasive and observable sample collection. Cannabinoids are usually the most prevalent analytes in illicit drug testing. Thus, scientific data are required to document sensitive and specific cannabinoids detection in OF if this technology is to be accepted for drug monitoring. The NIDA Chemistry and Drug Metabolism Section developed multiple clinical protocols to evaluate OF cannabinoid disposition during extended abstinence and following smoked cannabis, oral THC, and Sativex® (oromucosal spray containing approximately 1:1 THC and cannabidiol [CBD]). During extended abstinence, median THC and 11-nor-9-carboxy-THC (THCCOOH) detection windows were 24 h and 13 days, respectively, in chronic cannabis smokers; CBD and cannabinol (CBN) were detected only on admission. After a single smoked 6.8% THC cigarette, THC concentrations were 10-fold higher than CBD and CBN concentrations. Cannabinoids rapidly declined within 3 h; THCCOOH showed a more delayed elimination pattern. After single and multiple oral THC doses, parent cannabinoid concentrations reflected residual excretion from previous self-administered smoked cannabis, whereas THCCOOH concentrations were influenced by oral THC dose regimen. Conversely, parent cannabinoid concentrations were highly increased after Sativex, with CBD concentrations as high as THC's, distinguishing Sativex intake from smoked cannabis. Different cannabinoids had distinct elimination profiles and detection windows, which were influenced by administration route, dose, and prior drug intake history. Quantification of multiple cannabinoids is useful for establishing cutoff criteria for different drug testing programs; CBD and CBN may identify recent intake, while THCCOOH minimizes the potential for false positive results due to passive environmental exposure. Dry mouth, bloody OF samples, and OF cannabinoid stability may influence accurate cannabinoid quantification and should be controlled. OF/plasma cannabinoid ratios exhibited large inter-subject variability after smoking and during oral THC dosing, precluding blood concentration prediction from OF concentrations. There was a possible association among OF cannabinoids, smoking behavior, and cannabis tolerance/withdrawal/residual drug effects, suggesting that OF testing could be valuable for monitoring in cannabis dependence treatment programs. The present research provides an important body of data on cannabinoid OF disposition, improving interpretation of cannabinoid OF test results, formulating evidence-based drug control policy, and managing cannabinoid pharmacotherapy.