Saliva-Based THC Sensor Shows Promise as a Rapid Cannabis Screening Tool
A 2019 proof-of-concept study from the University of Texas at Dallas described a portable electrochemical sensor designed to detect tetrahydrocannabinol (THC) in saliva in less than a minute. The device could eventually support roadside screening, but it should not be confused with a validated test for driving impairment.
The research, published in Scientific Reports, addressed a practical problem in cannabis-impaired-driving enforcement. Blood testing can provide useful toxicological information, but collecting a sample is invasive, requires trained personnel and is rarely suitable as an immediate roadside procedure. Breath-based THC devices have also faced technical challenges because the amount of THC in exhaled breath can be extremely small.
Saliva, or oral fluid, offers a more accessible alternative. It can be collected with a swab, and THC is commonly detectable in oral fluid after cannabis use. However, the presence of THC in saliva does not establish that a person is impaired at that moment. As the Centers for Disease Control and Prevention explains, researchers still find it difficult to connect a specific THC concentration in an individual’s body with driving performance.
How the sensor works
The UT Dallas system uses an antibody-based sensing surface and electrochemical impedance spectroscopy. THC-related target molecules bind to the sensor’s capture chemistry, changing the electrical response measured across the electrode. An electronic reader then interprets that response to estimate the concentration in the saliva sample.
In laboratory testing, the sensor showed a detection limit of 100 picograms per milliliter and a working range extending from 100 picograms per milliliter to 100 nanograms per milliliter in human saliva. The researchers also tested saliva samples from human donors that had been prepared with known THC concentrations. Their classification model distinguished samples containing THC from THC-free samples with an area under the curve of 0.95.
Those results demonstrate analytical potential, not courtroom-ready evidence of impairment. The study was primarily an early validation of the sensing platform: it examined synthetic and prepared saliva samples and used a limited human-sample dataset. It did not establish a universal THC threshold for impaired driving, nor did it show that a particular reading predicts whether someone is unsafe to drive.
Why a positive result is not the same as impairment
Unlike alcohol, cannabis does not have a universally accepted concentration equivalent to a blood-alcohol limit. THC levels can change rapidly after use, and the relationship between measured THC, tolerance, route of administration, timing and actual driving ability is complex. The National Highway Traffic Safety Administration notes that evidence does not support a simple, direct dose-response relationship between blood THC concentration and driver impairment.
That limitation affects both blood and saliva testing. A rapid oral-fluid test could help an officer identify recent cannabis exposure and decide whether additional evaluation or confirmatory laboratory testing is warranted. It would be better understood as one component of a broader assessment—including observed driving behavior, field evaluation and, where appropriate, laboratory analysis—rather than as a stand-alone impairment meter.
From prototype to practical device
The researchers presented the original sensor as a possible foundation for a handheld roadside instrument. Its small sample requirement, short analysis time and compatibility with portable electronics could make it easier to deploy than conventional laboratory testing. The same general approach might also interest medical-cannabis researchers studying dose timing and individual responses.
Further development has followed. In a 2023 follow-up study in Analyst, the UT Dallas team described CannibiSenS, a field-deployable system designed to detect THC in saliva at a 5-nanograms-per-milliliter cutoff across a 0.1-to-100-nanograms-per-milliliter range. That work also examined selectivity against cannabidiol (CBD) and ethanol, while acknowledging that the human dataset remained limited.
Saliva-based THC sensors therefore represent a promising screening technology, not a replacement for impairment assessment. Their eventual value will depend on larger studies involving people who have used cannabis under controlled conditions, comparisons with validated toxicology methods and research linking test results with actual cognitive and driving performance.
