The rapid proliferation of new synthetic cannabinoid receptor agonists (SCRAs) has initiated considerable interest in the development of so-called "untargeted" screening strategies. One of these new screening technologies involves the activity-based detection of SCRAs. In this study, we evaluated whether (synthetic) cannabinoid activity can be detected in oral fluid (OF) and, if so, whether it correlates with SCRA concentrations. OF was collected at several time points in a placebo-controlled JWH-018 administration study. The outcome of the cell-based cannabinoid reporter system, which monitored the cannabinoid receptor activation, was compared to the quantitative data for JWH-018, obtained via a validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method. A total of 175 OF samples were collected and analyzed via both methods. The cannabinoid reporter assay correctly classified the vast majority of the samples as either negative (<0.25 ng/mL; 74/75 = 99%) or having low (0.25-1.5 ng/mL; 16/16 = 100% and 1.5-10 ng/mL; 37/41 = 90%), mid (10-100 ng/mL; 23/25 = 92%) or high (>100 ng/mL; 16/18 = 89%) JWH-018 concentrations. Passing-Bablok regression analysis yielded a good linear correlation, with no proportional difference between both methods (slope 0.97; 95% confidence interval 0.86-1.14) and only a small systematic difference. This is the first study to demonstrate the applicability of an untargeted, activity-based approach for SCRA detection in OF. Additionally, the outcome of the cannabinoid reporter assay was compared to the gold standard (LC-MS/MS), showing a good correlation between both methods, indicating that the cannabinoid reporter assay can be used for an estimation of drug concentrations.