THC saliva test accuracy: the short answer
A lab saliva test that uses mass spectrometry is accurate for recent THC use. Roadside saliva devices are not the equal of lab tests. In published studies, roadside devices detect THC with sensitivity between about 50% and 90%, based on the device and its cutoff level. Specificity is higher, above 90% for most devices. A positive roadside result is presumptive. It needs lab confirmation.
Saliva tests look for delta-9-THC itself. Urine tests look for THC-COOH, a metabolite that leaves the body over days. That difference explains the short detection window in oral fluid.
Detection window in saliva
- Single use, low dose: 12 to 24 hours.
- Regular use: 24 to 48 hours.
- Heavy chronic use: up to 72 hours in some studies.
- Urine, for comparison: 3 to 30 days. Hair: up to 90 days.
The National Institute on Drug Abuse lists a saliva window of about one to three days.
Cutoff levels decide the result
A cutoff is the THC concentration, in nanograms per milliliter (ng/mL), that a device calls positive. Federal oral fluid guidelines from SAMHSA set the screening cutoff at 4 ng/mL and the confirmation cutoff at 2 ng/mL. Roadside devices use cutoffs from 5 ng/mL to 25 ng/mL.
This matters. A device with a 25 ng/mL cutoff will miss a sample that a lab would confirm at 5 ng/mL.
Lab methods versus roadside devices
- Roadside: immunoassay screen on a swab sample. Read in 5 to 15 minutes. Presumptive result.
- Lab screen: immunoassay on a collected oral fluid sample. Quantified.
- Lab confirmation: liquid chromatography with tandem mass spectrometry (LC-MS/MS). This detects and measures THC and its metabolites.
False positives and false negatives
Oral fluid has fewer false positives than urine. Problems still occur.
- Cross-reactivity. Some immunoassay antibodies bind other cannabinoids, such as CBD and CBN, at high concentrations. Lab confirmation resolves this.
- Mouth contamination. Smoking right before collection leaves THC in the mouth. Swab results can read high with no systemic absorption.
- Dilution. Water, gum, or toothpaste before collection can lower THC in the sample and cause a false negative.
- Collection volume. Devices that collect small fluid volumes give less reliable results. Standard practice requires a fixed volume, often 1 mL.
What shifts the number
- Time since last use.
- Dose and frequency.
- Route. Edibles produce lower oral fluid THC than smoked flower in most studies.
- Device cutoff and sample volume.
Does a positive saliva test prove impairment?
No. Oral fluid THC shows that THC entered the body in the recent past. It does not measure coordination, reaction time, or driving skill. No saliva cutoff has been validated as an impairment limit. Blood, not saliva, is used for per se THC limits in the states that have them.