Marijuana is currently classified by the federal government as a Schedule I substance because it fulfills all requirements that define Schedule I. Nevertheless, the government has made two attempts to reclassify it to Schedule III, most recently through acting Attorney General Todd Blanche’s Order No. 6754–2026. The previous attempt was in 2024 under the Biden administration.
The Schedule I classification is reserved for those substances that have a high potential for abuse, no currently acceptable medical use, and lack an accepted safety for use under medical supervision.
As the founding president of DUID Victim Voices, my concern with rescheduling is its impact on highway safety: Driving Under the Influence of Drugs (DUID). Because of my concern, I was requested to testify at the DOJ hearing called to address the rescheduling of marijuana. At the hearing, I made the following points:
- The acting Attorney General’s 2026 order described many risks of marijuana use that could be exacerbated by rescheduling, but it completely ignored the risk of marijuana impairment on highway safety. DOJ’s earlier 2024 risk assessment included DEA and HHS evaluations that addressed the highway safety risk, recognizing that highway safety should be evaluated when considering rescheduling. However, both agencies’ assessments were so shallow as to be incompetent.
- There is a large and consistent body of experimental and epidemiological evidence proving that marijuana’s THC impairs driving and causes crashes — many of them fatal.
- Alcohol, likewise, poses severe risks to highway safety. All states ban driving while under the influence of alcohol (DUI) and substances like marijuana. It is possible, but difficult to prove, that a driver is under the influence of alcohol by observing and documenting well-known drunk driving symptoms. Fortunately, we have the “safety net” of DUI per se laws that also prohibit driving with excess alcohol in the driver’s blood or breath, typically, more than .08 g/dL or .08 g per 210 L of breath. DUI per se limits and other measures such as education, public pressure, increased enforcement, training, and better engineered roads and vehicles have dropped the traffic death rate per billion vehicle miles traveled by 2/3 in the last half century. That death rate has now plateaued and appears to be climbing.
- Note, however, that the chemistry and pharmacokinetics of marijuana’s THC make blood THC concentration a poor surrogate for impairment so there can be no scientifically valid per se safety net for marijuana like the one that exists for alcohol. Contrary to popular belief, our lack of a marijuana per se limit is not because we have not found the limit yet. It is because such a limit does not exist.
- Lacking marijuana per se limits or other objective measures of marijuana impairment, law enforcement struggles to enforce laws against drugged driving. The rate of convictions for drivers under the influence of marijuana lags far behind the conviction rate for drunk drivers. Past actions to loosen restrictions on marijuana access and use have coincided with dramatic increases in drugged driving and drugged driving deaths. Why should down-scheduling marijuana be any different?
The AAA Foundation for Traffic Safety, National Safety Council, National Highway Traffic Safety Administration (NHTSA), National Sheriff’s Association, International Association of Chiefs of Police and the National District Attorneys Association have all published policy statements acknowledging that there is no scientific basis for the adoption of THC per se laws for marijuana. Adopting scientifically invalid THC per se laws risks both convicting innocent drivers and exonerating impaired drivers. Seeking a nonexistent THC per se level is clearly not the answer.
Until validated objective technologies to measure, document, and prove a driver is impaired by marijuana or other drugs are available, any effort to aid the commercialization and use of drugs like marijuana is illogical, immoral, and irresponsible. No such technologies currently exist. Efforts to create such technologies based upon THC measurements are doomed to failure.