Police officers across the United States are using roadside tests to detect marijuana-impaired driving that are “not much better than a coin toss,” according to a scathing new analysis published in the Journal of Studies on Alcohol and Drugs.
The editorial argues that methods employed by thousands of specially trained officers rely on pseudoscience rather than validated scientific techniques, creating serious legal and public safety concerns.
With marijuana now legal in many states, the need for reliable impairment detection has become urgent. Unlike alcohol, which can be measured objectively with breathalyzers, marijuana presents unique challenges that have led law enforcement to rely on subjective evaluation methods that lack scientific foundation.
“The existing evidence suggests they’re ‘not much better than a coin toss,’” said William J. McNichol, an adjunct professor at Rutgers University Camden School of Law and author of the perspective piece.
The Drug Recognition Expert Problem
At the center of the controversy are Drug Recognition Experts (DREs)—more than 8,000 police officers nationwide who follow a standardized protocol claimed to detect drug impairment and identify specific substances. The process involves multiple steps including coordination tests, blood pressure checks, pulse measurements, and examining pupil size and eye movements.
However, McNichol argues this approach represents “police science”—techniques developed by officers for police work rather than evidence-based methods. The protocol includes questionable indicators such as squeezing drivers’ limbs to assess “muscle tone” and diagnosing psychiatric conditions like paranoia on the roadside.
A critical detail not emphasized in the press release: A 1998 controlled laboratory study found the DRE protocol produced false negative or false positive results 45.5% of the time. Even more troubling, when two groups of DREs evaluated the same subjects, they agreed on cannabis impairment only 69% of the time.
Scientific Challenges
The fundamental problem stems from attempting to apply alcohol-detection methods to marijuana without accounting for crucial differences between these substances:
- Alcohol is water-soluble; cannabinoids are fat-soluble and metabolized differently
- THC blood levels don’t correlate with impairment levels
- Heavy users develop tolerance, complicating standardized testing
- Multiple cannabinoids affect impairment, not just THC
- Some cannabinoids persist long after impairment ends
Legal and Social Consequences
The stakes extend far beyond academic debate. DUI convictions carry severe penalties including license suspension, job loss, and lasting criminal records. McNichol warns that relying on “untrustworthy evidence undermines the credibility” of impairment laws and “contains within itself the seeds of its own destruction.”
Several courts have begun excluding DRE testimony, with Maryland, Michigan, and Rhode Island rejecting it entirely. New Jersey allows DREs to testify only that behavior was “consistent with” marijuana use, while Minnesota forbids officers from calling themselves “experts.”
