As marijuana use grows, effect on road safety remains a blind spot

In Colorado, where the United States’ experiment with legal recreational marijuana began a little more than a decade ago, a team of federal scientists has been paying regular cannabis users to get stoned.

This unconventional line of research — which includes vans outfitted with hippie tapestries and a sleek car simulator — seeks to tackle what road safety experts regard as a serious blind spot as marijuana use grows nationally.

Law enforcement officials lack tools to detect cannabis-impaired driving as reliably as they can identify people who get behind the wheel drunk.

Only a few states routinely test the blood of drivers involved in serious accidents for marijuana, and as a result, little is known about how cannabis use is affecting road safety. Police officers generally need a warrant to compel a driver suspected of being impaired to provide a blood sample.

Even when blood samples are analyzed, tests cannot reliably establish whether a person last used marijuana hours before the accident or several days prior, making the tests an imprecise gauge of impairment.

Complicating matters, state laws on cannabis-impaired driving are inconsistent and confusing, which has made them difficult for the police to enforce and for motorists to understand.

“We’re kind of painting the plane as we fly it when it comes to cannabis liberalization,” said Jake Nelson, the director of traffic safety advocacy and research at AAA, the automobile drivers group, which opposes the legalization of recreational cannabis. “Public health and safety has been more of an afterthought.”

National data on the effects of marijuana use on road safety is spotty, but studies have shown that cannabis can impede a driver’s ability to respond quickly to an obstacle and judge distance accurately.

In Colorado, the federal scientists are pursuing a pair of studies they hope will help policymakers craft sensible, enforceable standards. For starters, the scientists are working to design portable Breathalyzers that come closer to establishing how recently a driver used marijuana.

Separately, by observing scores of stoned drivers using a car simulator, they hope to more clearly understand how, and at what levels, cannabis impairs motor skills and reflexes of both habitual and occasional users.

“Policymakers need data on which to base their policies,” said Tara M. Lovestead, a chemical engineer at the National Institute of Standards and Technology, the federal agency overseeing the studies. “And there isn’t much data here.”

In recent decades, marijuana became far more accessible as 39 states sanctioned its medical use and 24 states legalized recreational weed. Roughly 15% of American adults reported that they smoke marijuana, according to a Gallup survey from 2023 and 2024, compared with 7% of people in 2013.

And a substantial number of Americans acknowledge that they are driving after consuming marijuana. An estimated 14.8 million motorists said that they have operated a vehicle within an hour of using marijuana, according to an AAA survey from 2019. A survey that Virginia’s cannabis control agency conducted last year showed that 22% of drivers in that state admitted to having driven stoned a few times a year or more.

Lawmakers in at least 10 states have prohibited driving with any level of THC, the intoxicating component of cannabis, in a motorist’s system. Legislators in several other states, including Montana, Illinois, Ohio, Washington and Nevada, have passed marijuana driving laws modeled after drunk-driving laws, in which the national standard has become a limit of .08 blood-alcohol concentration. Those states have adopted a range of such limits, measured by the amount of THC in a person’s blood.

But experts say the approach used with drunk-driving does not easily lend itself to marijuana.

There is no consensus on what level of cannabis use may be safe for driving. That is partly because unlike alcohol, which impairs people in predictable, well-documented ways, the effects of cannabis vary widely based on whether it’s smoked, eaten or drunk.

People who smoke get high within minutes but effects wear off fairly quickly. Edibles generally are felt within an hour or so, and tend to induce altered states that last longer. Complicating matters, people who use cannabis often tend to become less impaired over time, experts say.

Colorado and Washington, which became the first states to legalize recreational marijuana about a decade ago, have taken steps to study cannabis and road safety, though recent statistics from those states show no consistent trend.

Both states saw an increase in the number of fatal accidents between 2018 and 2022 in which a driver’s blood contained marijuana, but officials say they cannot be sure how many of the motorists were impaired at the time of the accident because THC can often be detected in the blood long after the psychoactive effects wear off.

Pam Fischer, an official at the Governors Highway Safety Association, a nonprofit representing state highway safety agencies, said many states have done relatively little to study and mitigate the implications for road safety brought about by legalizing marijuana.

“There was a lot of emphasis on the tax revenue that was going to come into the state,” said Fischer, a former director of the New Jersey Division of Highway Traffic Safety. “But we need to think about the societal harm that can happen if we aren’t careful and thoughtful.”

Mary Kazmark, whose daughter, Krystal, was killed in a crash in California in 2020, said she did not have strong views about legalizing marijuana until the accident. Her daughter’s boyfriend, Joshua Daugherty, who was driving, told investigators he had smoked marijuana about eight hours before the crash, according to a California Highway Patrol accident report. Daugherty told officials that he had lost control of his car while trying to avoid hitting an animal on the road and asserted that his frequent use of marijuana meant that it did not impair him, according to the report.

Kazmark said she was disappointed and surprised when Daugherty was charged with vehicular manslaughter without gross negligence, a misdemeanor. California is among states that outlaw impaired driving, but have not set a specific limit on cannabis use and driving.

Kazmark said she was deeply frustrated that the driver served only a few months behind bars after pleading guilty to the misdemeanor charge.

“I just wish that they could find a way to measure the intoxication and then hold people accountable,” she said. “If they had that in place, more people would think twice before lighting up.”

Greg Hayes, a prosecutor in Sacramento, California, said prosecutions of impaired drivers who have used cannabis are rare because investigators lack tools to prove impairment. Another challenge, he said, is a widespread perception by people who consume marijuana that they can safely drive afterward — a claim that is sharply disputed by federal safety officials.

“You can smell marijuana off the jurors in the courthouse,” Hayes said. “So you’re going in front of jurors, some of whom, that very morning when they came to do jury selection, smoked marijuana and drove in.”

Ashley Brooks-Russell, a public health professor at the University of Colorado who is running the federally funded driving simulator study, said alcohol and cannabis alter driving performance in different ways. Drunken drivers tend to speed, tailgate and act more aggressively than sober ones. Drivers who have used cannabis, on the other hand, tend to slow down, she said.

Brooks-Russell said policies on cannabis and driving should not be modeled after drunk-driving laws.

“I think there’s a wishful thinking that it should be like alcohol,” she said. “But it’s not.”

© 2025 The New York Times Company