Medical Marijuana Just Got Easier to Research. Here’s What We Already Know — and Don’t.
Rescheduling marijuana to Schedule III is a policy choice, not a scientific finding. The research supports a short list of uses, patients are buying products nothing like what was tested.
Schedule I was scientifically wrong, and rescheduling fixes that.
Strong evidence for four conditions, weak or missing for the rest.
Dispensary products look nothing like what was tested in trials.
One in three medical users develops problem use within a year.
Schedule I was scientifically wrong, and rescheduling fixes that.
On April 23, 2026, the Department of Justice announced it was moving marijuana from Schedule I to Schedule III under federal drug law. Schedule I is for drugs the government says have no medical use. Schedule III is for drugs with medical use and a moderate risk of dependence, the same category as testosterone and ketamine.
The change followed a December 2025 executive order directing the attorney general to lower marijuana’s classification. The White House said the goal was to expand research and protect patients in states that had already legalized medical cannabis.
Schedule I was scientifically wrong. Marijuana does have medical uses. Keeping it in the same category as heroin made the whole scheduling system look absurd and created real barriers to clinical research. Rescheduling removes those barriers. But it does not, on its own, fill the gap in evidence that matters most to patients.
Strong evidence for four conditions, weak or missing for the rest.
In November 2025, JAMA published the largest review of cannabis research to date, covering more than 2,500 studies.
Strong evidence supports marijuana for a short list of conditions: long-term nerve pain, nausea and vomiting from chemotherapy, muscle stiffness from multiple sclerosis, and certain childhood seizure disorders like Dravet syndrome and Lennox-Gastaut syndrome. For sleep problems, anxiety, PTSD, and Alzheimer’s disease, the evidence is weak or missing.
Even where the evidence is solid, the benefits are small. For chronic pain, doctors would need to treat 11 patients before one gets meaningful relief beyond a placebo. On a standard 10-point pain scale, the average improvement was half a point.
The FDA has approved exactly three marijuana-based drugs: cannabidiol (Epidiolex) for childhood seizures, dronabinol (Marinol) for HIV/AIDS-related weight loss and chemotherapy nausea, and nabilone (Cesamet) for chemotherapy nausea. These are the uses where the benefits clearly outweigh the risks. They cover a small fraction of the people who use medical marijuana.
Patients are running ahead of the science.
More than 4.1 million Americans were enrolled in state medical marijuana programs in 2022, up 4.5-fold from 2016. The top reason is chronic pain, at about 48%. Anxiety is 14%. PTSD is 13%.
The share of patients using marijuana for conditions with strong scientific backing has been shrinking. In 2016, 85% of medical marijuana patients had evidence-supported conditions. By 2022, that fell to 54%. States kept adding qualifying conditions. The research did not keep pace.
This is exactly why rescheduling for research matters. We need rigorous trials on the conditions for which millions of patients are already using marijuana. Right now, those patients are running ahead of the science.
Dispensary products look nothing like what was tested in trials.
The JAMA review raised a second concern. Clinical trials that showed benefit used products with THC levels below 10%. THC is the chemical in marijuana that produces a high and is linked to most side effects. Dispensary products labeled “medicinal” average 19% THC. “Recreational” products average 22%.
That gap carries consequences. High-potency marijuana, above 10% THC, carries measurably higher risks. Users report psychotic-like experiences at nearly twice the rate of those using lower-potency products (12% vs. 7%) and anxiety disorders at a higher rate (19% vs. 12%). Clinical guidelines recommend against smoking or using high-potency marijuana for medical purposes.
What patients buy at a dispensary looks nothing like what was tested in the studies.
One in three medical users develops problem use within a year.
Cannabis use disorder, problem use that meets clinical criteria for addiction, affects 29% of medical users within six to twelve months. One in three patients who smoke marijuana 20 or more days a month develops cannabis hyperemesis syndrome: severe, repeated nausea and vomiting, which is one of the conditions marijuana is supposed to treat.
Daily smoking doubles the risk of heart disease (2% vs. 1%) and nearly triples the risk of stroke (3% vs. 1%), even after accounting for tobacco. Marijuana also changes how the body processes common drugs, including blood thinners, seizure medications, and antidepressants.
The patients most likely to use medical marijuana are the most vulnerable to its harms. In health records, 61% of medical marijuana users had conditions where it posed added risks. Mental health disorders showed up in 36%. Substance use disorders in 22%.
Support for marijuana legalization is falling, not rising.
Gallup polling from October 2025 found 64% of Americans view marijuana positively, down from 70% a year earlier. Republican support fell from 53% to 40%. YouGov polling from April 2026 found 53% support legalization, down from 60% in 2022. That softening tracks with growing visibility of cannabis-related harms — emergency room visits, psychotic episodes, hyperemesis — that were less common when marijuana was harder to get.
Rescheduling was step one. Here's what step two looks like.
Rescheduling was the right call. What it should unlock, and what will determine whether it actually helps patients, is a wave of rigorous, well-designed clinical trials on the conditions people are already using marijuana for. Trials that test products at the potency patients actually buy, not the low-dose formulations from a decade ago.
Beyond research, the clinical system needs structure. About 29,000 doctors authorized medical marijuana in 2022. More than half were in general practice. There is no standard training. There is no standard list of approved products. There is no system to make sure what patients buy at a dispensary has anything to do with what was studied in the trials.
The federal government just told patients that medical marijuana is real medicine. Now it has to build a system that treats it like one.