Air India pilot's reported cannabis test: What it can and can't tell investigators

Reports that an Air India pilot tested positive for cannabis after a flight sound like the sort of finding that should make an aviation investigator sit up.

But the reported result has not been officially confirmed, and even a confirmed positive would not necessarily be the smoking gun it might appear to be.

On 4 August, an Air India Airbus A320 flying from Phuket to Delhi suddenly lost about 91m (300ft) of altitude, injuring 13 passengers and four crew members. The aircraft, carrying 137 passengers and eight crew, landed safely in Delhi. The incident is under investigation.

Both pilots underwent routine post-flight screening for psychoactive substances, India's aviation ministry said. It said the captain's initial test produced a result requiring "confirmatory testing" , external and that both pilots had been removed from flying duties pending the inquiry.

Earlier this week, a number of media reports, citing sources, reported that the captain subsequently tested positive for marijuana in a confirmatory laboratory test. Air India has not confirmed the result. The reports have triggered a storm over pilot drug testing and aviation safety in India.

There is, so far, no official finding linking the captain's alleged drug-test result to the sudden loss of altitude. Investigators are examining the incident more broadly, including technical, operational, medical and human factors, and it remains unclear what caused the aircraft to plunge.

The obvious question is whether cannabis had anything to do with it.

The less obvious one is harder: what does a positive cannabis test actually tell us about a pilot's state of mind several hours earlier?

Quite a lot is known about how cannabis affects performance of complex tasks. Much less is known about how to turn a positive test into a reliable timestamp for impairment.

Some of the most directly relevant evidence comes from a rather unusual source: pilots.

In a 1976 experiment , external , 10 pilots smoked a modest dose of marijuana and then flew a simulator. Their performance deteriorated, with errors in altitude, heading and radio navigation. The effects were evident for at least two hours and generally disappeared within four to six hours, the study found.

A decade later, researchers returned to the question of what happens after the obvious "high" has worn off. , external

Ten experienced private pilots were tested on a flight simulator 24 hours after smoking cannabis. They showed significant deterioration in several measures, including elevator and aileron control, landing accuracy and vertical and lateral deviation on approach. Most strikingly, the pilots did not realise that their performance had deteriorated, the study found.

A 1991 follow-up , external produced a similar warning.

Nine active pilots flew a simulator after smoking a cigarette containing 20mg of THC, or tetrahydrocannabinol - the main psychoactive compound in cannabis and the chemical responsible for the "high" and many of the drug's effects on the brain and motor skills.

Performance was impaired when checked 15 minutes, four hours, eight hours and 24 hours later. Seven of the nine pilots showed some impairment at 24 hours, but only one was aware of the drug's effects, the study found.

These studies are small, old and were conducted with aircraft simulators rather than modern airliners. They cannot tell us what happened in the Air India cockpit. But experts believe they likely establish something useful: cannabis can interfere with the sort of fine motor control, attention and complex human-machine interaction that flying demands.

A review , external by aviation-medicine specialist David Newman for Australia's Transport Safety Bureau reached a similar conclusion.

Cannabis affects behaviour, cognition and psychomotor performance in a dose-dependent way, and complex tasks such as flying are particularly sensitive. The review also found evidence of residual effects lasting up to 24 hours.

The science has become more complicated, however, as cannabis use has become more widespread and the potency of cannabis products has increased in many parts of the world.

A 2022 randomised clinical trial , external led by Thomas Marcotte at the University of California San Diego tested 191 regular cannabis users.

Those who smoked THC performed worse on a simulated driving task, but the researchers found that impairment could not be inferred from the THC content of the cannabis, the users' previous experience or their blood THC concentration.

The drug initially impaired performance, but after about four-and-a-half hours, those who had taken it were performing about as well as people who had not received the active drug.

That finding matters enormously for aviation testing.

A positive test can show that THC is present, but not necessarily when cannabis was consumed or whether the person was impaired at the time. The relationship varies with the dose, how it was consumed, frequency of use and individual biology, experts say.

This is one reason cannabis presents regulators with a problem that alcohol does not.

"Unlike with alcohol where there's a more robust relationship between alcohol concentrations in blood and impairment, the association is much more complex with THC. THC rapidly leaves the bloodstream after smoking, and there is limited to no correlation between blood THC concentrations and impairment," Marcotte told the BBC.

"In people who use at least occasionally, THC can be detectable in blood for days and sometimes weeks after last use, particularly in more frequent users. This is long after any acute impairment has subsided."

Studies show that impairment after smoking cannabis typically lasts three to five hours. When cannabis is ingested, such as in an edible, impairment lasts longer - typically six to eight hours, and sometimes more.

After edible use, the metabolite 11-OH-THC (11-hydroxy-THC) is often more closely associated with impaired cognitive function than THC itself, which may be present at lower concentrations, says Marcotte.

Not surprisingly, there is no equivalent universally accepted cannabis test that can say: this person is currently too impaired to fly.

Canada has responded by taking a deliberately blunt approach. After legalising recreational cannabis in 2018, Transport Canada introduced a rule , external prohibiting flight crews and flight controllers from consuming cannabis for at least 28 days before duty.

The regulator says the rule was intended to address the uncertainties surrounding cannabis impairment and detection; and airlines can impose stricter rules.

The American aviation experience offers another warning against reading too much into a positive test.

A 2026 US National Transportation Safety Board (NTSB) study , external of pilots killed in US civil aviation accidents between 2018 and 2022 found that 52.8% had at least one drug detected, while 27.7% had two or more. THC was the main driver of the increase in illicit-drug detections. But the NTSB stressed that the study was about drug presence, not whether pilots were impaired.

Clearly, finding cannabis in someone's system does not, by itself, establish that it caused an incident or accident.

But there are cases where investigators have gone further. In a 2011 Canadian crash, a Cessna 208B pilot had 50.1 nanograms per millilitre (ng/ml) of THC in his blood. The Transportation Safety Board concluded , external that the amount of psychoactive cannabinoids was sufficient to impair cognitive processes and likely affected the pilot's planning and conduct of the flight.

In a 2017 New Zealand helicopter crash , external , THC was detected in the pilot's blood after cannabis use two days earlier. A forensic toxicologist concluded that he was unlikely to have been under its influence at the time of the accident, and investigators found cannabis very unlikely to have contributed.

The lesson from these cases is not that a positive cannabis test means nothing. It is that its meaning depends on the concentration, timing, circumstances and other evidence available to investigators

In the Air India case, investigators will need to establish much more than the alleged fact that the captain's confirmatory test was positive.

They will need to know what sample was tested, whether it contained THC itself or a metabolite - a substance produced when the body breaks THC down - at what concentration, and what that can say about the timing of use.

Air India is to immediately test all its pilots for banned substances after the incident, expanding on existing rules that require at least 10% of pilots to undergo random screening.

Urine, blood and even saliva can show that someone was exposed to THC, but they cannot reliably establish when it was used or whether the person was impaired at a particular point several hours earlier, according to Marcotte. Saliva is more closely linked to recent use, but can still detect THC many hours - and potentially a day or two - after consumption, he says.

Experts believe the question is not simply: Was the pilot using marijuana?

It is threefold: Was cannabis consumed? Was the pilot impaired while flying? And did that impairment contribute to the aircraft's sudden loss of altitude?

The first may soon be answered. The other two are considerably harder.