Does CBD show up on a saliva drug test? What France checks

French drug tests target Δ⁹-THC, the psychoactive molecule in cannabis, and THC-COOH, an inactive metabolite the liver produces when it breaks THC down. The roadside kit screens for four drug families: cannabinoids, amphetamines, cocaine and opiates. For cannabinoids, the decree targets only THC and its metabolite: cannabidiol does not appear in it, and the French Highway Code sets no CBD threshold.

The cut-offs come from the decree (arrêté) of 13 December 2016 on drug testing of drivers. Each one is a detection limit, the level at which the test must react. None of them is a tolerance limit.

The roadside kit is a rapid test: the decree requires it to detect THC from 15 ng/ml of saliva, and it does not measure how much. The offence of driving after drug use (article L235-1) is only established by the laboratory analysis, done by chromatography coupled with mass spectrometry, which detects THC from 1 ng/ml of saliva, a cut-off fifteen times lower, or from 0.5 ng/ml of blood.

The Highway Code targets a driver for whom "it results from a blood or saliva analysis" that they used drugs: the kit alone is not enough. Since 20 August 2026, article L237-1 also punishes driving "having manifestly used" drugs with the same 3 years and €9,000, without requiring any analysis. The laboratory must keep samples at −20 °C for at least one year and undergo external quality control (decree, article 13).

Grenoble University Hospital analysed 3,051 confirmation samples from drivers who screened positive, collected between 2018 and 2021 (Willeman et al., 2023): 98.2% of kit positives for cannabis were confirmed, against 22.6% for amphetamines and 17.3% for opiates, and 22.3% of samples showed at least one disagreement between kit and laboratory, all substances combined. The study says nothing about drivers who screened negative, whose samples were not sent. Under controlled conditions, against an analysis at a 10 ng/ml cut-off, 5 to 10% of the two kits' results were false positives and 9 to 16% false negatives, and neither met the recommended 80% for sensitivity, specificity and accuracy all at once (2019 Australian study).

A negative screen therefore does not prove that no THC is present. And each sample answers its own question: according to Drogues Info Service, saliva and blood mostly point to recent use, urine to recent or older use, and hair to long-term use, covered in the hair test.

The 0.30% threshold applies to the sale of hemp and has nothing to do with driving. A compliant full-spectrum product contains small but real amounts of THC: a flower at 0.25% THC carries 2.5 mg per gram. With frequent use, that THC eventually crosses the cut-offs above. Our article on the legal THC level explains the sales threshold.

A team at McLean Hospital, which is affiliated with Harvard, put numbers on this. Their study, published in 2021 in JAMA Psychiatry, followed 14 adults taking a full-spectrum hemp oil every day. Each took about 35 mg of CBD and 0.8 mg of THC daily, from an oil assayed at 0.02% THC. After four weeks, 7 of the 14 participants had a urine test positive for THC-COOH, with an oil holding fifteen times less THC than the French ceiling.

Taking CBD at the same time does not mask THC: in the Australian study cited above, the THC curve in saliva was the same with or without an equal amount of CBD. Drogues Info Service, the public drug information service run by Santé publique France, makes the same point in a 2025 note on CBD and saliva tests: a driver who tests positive after CBD took a product that contained THC.

Can CBD turn into THC in the body?

No, according to human data. The conversion seen in simulated gastric fluid has not been found in vivo, even after high oral doses of CBD (Nahler et al., 2017). French toxicologists consider that this conversion appears, under certain conditions, to be an analytical artefact, and point out that a CBD product with up to 0.3% THC, sometimes much more for products sold online, can make both screening and confirmation positive (Alvarez et al., Thérapie, 2023). In the body, the THC therefore comes from the product.

A few hours after one-off use, longer in a daily user, and no length of time guarantees a negative test. The window depends on how often you use, how the product is taken and above all on the test's cut-off: published studies find THC in saliva for a few hours to a little over three days, at laboratory cut-offs of 0.2 to 2 ng/ml.

The table of positivity windows published by Drogues Info Service (June 2025) gives these orders of magnitude:

• Saliva: 6 to 8 hours after occasional use, up to 24 hours with regular use, up to 8 days with heavy daily use.
• Blood: THC detectable 2 to 8 hours after occasional use, THC-COOH up to 72 hours; over a month with heavy use.
• Urine: 3 to 5 days after occasional use, 30 to 70 days with regular use.

The table calls occasional at least one use in the year, regular more than one use a week (10 or more in the month), daily one use every day, and heavy daily use several uses a day over a long period. These windows vary widely from one person to another.

These cut-offs of 0.2 to 2 ng/ml (1 µg/l equals 1 ng/ml) are laboratory cut-offs, and none of these windows holds for a given person: Drogues Info Service points out that no method can predict precisely how long a test stays positive, as it depends on many factors.

No official source publishes a window specific to CBD products, and the test does not look for CBD itself. How long CBD stays in the body and the 2025 trial are covered in how long CBD stays in the body, and the detection-window tool gives an order of magnitude, with no guarantee. No waiting time guarantees a negative test: do not drive after a product that contains THC.

Article L235-1 of the French Highway Code punishes driving after drug use with 3 years in prison and a €9,000 fine, penalties raised by the law of 9 July 2025, and with 5 years and €15,000 when alcohol is also involved. The offence also costs 6 licence points and can lead to a licence suspension of up to 5 years, according to the « Drogue au volant » page on service-public.fr.

The law sets no minimum amount. The Court of Cassation confirmed this on 21 June 2023 (appeal no. 22-85.530): the authorisation to sell hemp derivatives below 0.30% THC "has no bearing on the offence of driving after drug use". Having bought a legal product is therefore not a defence. The penalties are detailed in CBD and driving: what the law of 9 July 2025 tightens.

If the kit is positive, a second saliva sample goes to a laboratory. That is the moment to say you want to reserve a counter-analysis (contre-expertise): a blood sample is then taken. The request itself is filed later, within 5 days of being notified of the result.

Screening is mandatory after an accident causing death or injury, possible after an accident with material damage only, a presumed traffic offence or a plausible reason to suspect drug use, and even with no reason at all, on the prosecutor's requisition or on the initiative of judicial police officers (article L235-2). The check then runs in this order:

1. The kit, whose result shows within a few minutes, according to Drogues Info Service. If it is negative, the check ends there.
2. If it is positive, you collect a second saliva sample yourself, under the supervision of a judicial police officer (decree, art. 7).
3. The officer asks whether you want to reserve a counter-analysis or a search for psychoactive medicines. If you say yes, a doctor, nurse or biologist called in takes a blood sample as soon as possible, and a doctor carries out a clinical examination (article R235-6).
4. The laboratory analyses the sample and keeps a tube of blood for a possible counter-analysis (article R235-9).
5. Within 5 days of being notified of the results, you ask for the counter-analysis from the prosecutor, the investigating judge or the court named on the notification; a different laboratory carries it out. After a saliva analysis, it requires the reservation made at step 3 (article R235-11).

Refusing the checks is a separate offence, punished by 3 years in prison, a €9,000 fine and the loss of 6 points (article L235-3). Refusing the kit stops nothing: a saliva or blood sample is then taken.

In 2022, 776,000 drug screenings, 19% more than in 2021, led to 127,000 offences, according to a French Senate report. For a police check on foot, see police checks and CBD.

Cannabis screening tests are freely sold, in pharmacies as well as online, according to Drogues Info Service. Their result is only indicative and has no place in the Highway Code procedure; a negative self-test does not prove that no THC is present.

The leaflet gives the test's cut-off, the reading time and how to spot an invalid test. Whatever that cut-off, a self-test does not predict a later check: THC levels in saliva change from hour to hour and kits err in both directions. A negative self-test therefore does not clear you to drive: after a product that contains THC, do not drive. It does not detect CBD and says nothing about a product's THC content.

An employer cannot test just anyone. Workplace saliva testing is limited to jobs where a mistake endangers the employee or others, and it must be provided for in the company's internal rules. The exact conditions and employee rights are in what an employer can legally test.

Can you refuse a saliva test at work?

On the road, refusing the checks is an offence. At work, the Highway Code does not apply: the test must be provided for in the internal rules for a safety-sensitive job, and the consequences of a refusal, which are disciplinary, are assessed within that framework.

For athletes, the World Anti-Doping Agency removed CBD from its prohibited list in 2018. THC remains banned in competition above 150 ng/ml of urine. An athlete taking a full-spectrum oil faces the same risk as a driver; the full picture is in CBD and doping: the 2026 WADA list.

The risk differs across three product families:

What reduces the risk:

• Read the Δ⁹-THC line on the batch certificate of analysis: "ND" (not detected) or "< LOQ" (below the limit of quantification) is not the same as "0.2%". Our guide to reading a certificate of analysis goes through each line, the laboratory page describes the measurement methods, and THC-free CBD oil explains what the label covers.
• If you drive or are tested at work, choose an isolate. What "99%" guarantees is explained in CBD isolate: what 99% means, and the formulation differences in full spectrum or isolate.
• Do not drive after using a product that contains THC: no waiting time guarantees a negative test.

What does not work: quick "detox" products, litres of water before a test, and a certificate kept in the glovebox. It shows good faith, but the court looks at the analysis. Common beliefs about clearing THC are reviewed in clearing THC: myths and facts.

If your cannabis use, or a relative's, worries you, Drogues Info Service answers on 0 800 23 13 13, from 8 am to 2 am, 7 days a week, anonymously and free of charge, and its useful addresses lead to youth consultation services (CJC) and addiction treatment centres (CSAPA).