For most of living memory, an American scientist who wanted to study cannabis needed a federal license that took years to get, a locked storage system, and a supply of research plants from a single farm in Mississippi that grew nothing resembling what people actually used. Heroin was easier to research. Then, in December 2025, a presidential order titled "Increasing Medical Marijuana and Cannabidiol Research" set off the biggest change in federal cannabis law in half a century. This is the story of how the science got locked out, what it managed to prove anyway, and what the 2026 changes mean for anyone who has ever wondered what cannabis really does.
Looking for Cannabis Studies Near You?
Browse open Cannabis clinical trials! We'll handle the follow-up.
How a medicine became impossible to study
Cannabis was in American pharmacies until the 1930s, sold as a tincture for pain, spasms, and sleep. The 1937 Marihuana Tax Act effectively criminalized it, and research mostly stopped. Then came the Controlled Substances Act of 1970, which sorted drugs into five schedules. Schedule I was reserved for substances with a high potential for abuse and no accepted medical use. Cannabis was placed there provisionally, alongside heroin and LSD, pending a commission's review. The commission recommended decriminalizing it in 1972. The recommendation was ignored, and cannabis stayed in Schedule I for the next 55 years.
The consequences for research were mechanical, not philosophical. Schedule I status meant every researcher needed a separate DEA registration to possess the substance, a process that routinely took a year or more. It meant institutional review boards treated cannabis studies as high risk. It meant funding flowed overwhelmingly to studying harms rather than benefits, because the National Institute on Drug Abuse, whose mission is drug abuse, was the main federal funder. And it meant one extraordinary bottleneck: under the 1961 international drug treaty and the 1970 law, the federal government had to be the single source of research cannabis, and from 1968 until 2021 it designated exactly one grower, the University of Mississippi, under contract to NIDA.
The one farm problem
For 53 years, every legal research cannabis plant in the United States came from Oxford, Mississippi. Scientists who received it described dried, low potency material shipped as rolled cigarettes, with THC content far below what dispensaries sold, and no concentrates, edibles, vapes, or high CBD strains available at all. A study of the government supply published in 2017 found it was genetically closer to hemp than to the cannabis sold in legal states. Researchers were, in effect, studying a different product from the one Americans used.
The DEA announced in 2016 that it would license additional growers. Then nothing happened for years. Applicants sued; one, the Scottsdale Research Institute, went to a federal appeals court in 2019 to force the agency to respond to 33 pending applications. The first agreements with new growers came in May 2021. By 2025, seven licensed growers existed, and NIDA quietly ended its half century contract with Mississippi. Meanwhile, Congress passed the Medical Marijuana and Cannabidiol Research Expansion Act in December 2022, the first standalone federal cannabis law ever, which streamlined researcher registration and ordered the government to report on cannabis's therapeutic potential. Each step helped. None of them changed Schedule I.
What the science proved anyway
Despite all of this, evidence accumulated, much of it from other countries and from synthetic cannabinoids that were easier to study. In 2017 the National Academies of Sciences, Engineering, and Medicine reviewed more than 10,000 studies and drew the most careful map the field has. They found conclusive or substantial evidence for exactly three things: cannabis or cannabinoids reduce chronic pain in adults, oral cannabinoids reduce chemotherapy induced nausea and vomiting, and cannabinoids improve patient reported spasticity in multiple sclerosis. For most other claimed uses, from anxiety to PTSD to glaucoma, they found limited or insufficient evidence, which is a statement about missing studies as much as about the plant.
The FDA's record tells the same story. Two synthetic THC medicines, dronabinol and nabilone, have been approved since the 1980s for chemotherapy nausea and, for dronabinol, appetite loss in AIDS. And in June 2018 the FDA approved Epidiolex, a purified CBD extract, for two severe childhood epilepsy syndromes, the first and still the only plant derived cannabinoid medicine ever approved in the US. It passed because a company ran the rigorous trials Schedule I made so hard: randomized, placebo controlled studies that showed seizures fell by roughly 40 percent compared with about 15 to 20 percent on placebo. That is what cannabis research looks like when the system allows it. Everything else Americans believe about medical cannabis, good or bad, rests on thinner ground than it should.
What changed in 2026
In May 2024, after a review by the Department of Health and Human Services concluded that cannabis has accepted medical uses and a lower abuse potential than other Schedule I drugs, the Justice Department proposed moving it to Schedule III, the tier of medicines like ketamine and codeine combinations. More than 42,000 public comments came in. A hearing was scheduled, then stalled through 2024 and 2025 on procedural disputes.
On December 18, 2025, President Trump signed Executive Order 14370, "Increasing Medical Marijuana and Cannabidiol Research," directing the Attorney General to finish the rescheduling as fast as the law allows. On April 23, 2026, the Justice Department and DEA issued a final order that did two things. It immediately moved FDA approved cannabis products and cannabis sold under state medical marijuana licenses from Schedule I to Schedule III, the most consequential change to federal cannabis law since 1970. And it set an expedited hearing, which began on June 29, 2026, to consider moving all cannabis to Schedule III. As of this writing the hearing process is still underway, and cannabis sold for adult use remains in Schedule I.
What Schedule III changes for research is concrete. Researchers no longer need the special Schedule I registration for medical cannabis; the standard registration used for thousands of prescription medicines applies. Institutional review boards can treat cannabis studies like other drug studies. Universities can more easily obtain state licensed products that reflect what people actually use, rather than the government supply. Funding can flow to studying benefits, not only harms. And companies can run FDA trials with far less friction, which is the path Epidiolex took and almost nothing else has. Experts caution that none of this happens overnight: registrations still take time, the hearing could reshape the details, and cannabis remains federally illegal outside these channels. But the structural reason so little was known has been removed for medical cannabis, and may be removed entirely.
What researchers are studying right now
The research that was possible even under Schedule I is already recruiting, and it spans a much wider range than most people expect. Much of it involves giving cannabis, THC, or CBD to participants under medical supervision, because the basic questions about modern high potency products have never been answered under controlled conditions.
What cannabis actually does. Experimental studies in adults who already use cannabis measure its effects on sleep, circadian rhythms, and light sensitivity, on memory, attention, and brain waves, and on perception and stress. Participants receive cannabis supplied by the study in a lab setting. These are not treatment studies; they are the missing baseline.
CBD as medicine. Trials test cannabidiol for bipolar depression, for sleep, anxiety, pain, and mood in older adults who use cannabis, for recovery from opioid use disorder, and for people at risk for Alzheimer's disease. CBD is sold everywhere and studied almost nowhere; these trials are the exceptions.
Treating the problem side. About 3 in 10 people who use cannabis meet criteria for a use disorder, and there is no approved medication. Trials are testing psilocybin assisted treatment, brain stimulation, CBD itself in young adults who use cannabis, and whether repeated CBD changes cannabis use. Our cannabis use disorder page lists these.
The interactions. Studies of cannabis with alcohol, with tobacco, and in people with schizophrenia address how cannabis behaves in the way people actually use it, alongside other substances.
Why cannabis research will still take time
The product problem. "Cannabis" is not one thing. THC content, CBD ratio, terpenes, and delivery method (smoked, vaped, eaten) all change effects. A trial has to specify a product precisely, and standardized products that match the market are only now becoming available to researchers.
The blinding problem. Participants usually know whether they received THC, because they feel it. That makes placebo controlled trials, the gold standard, harder to run honestly, and researchers have developed low dose comparison arms and other designs to cope.
The funding lag. Decades of funding aimed at harms produced a research workforce, infrastructure, and literature skewed in that direction. Rebalancing toward therapeutic questions takes years of new grants and new trials.
The legal patchwork. Forty seven states and Washington DC allow some medical use, each with different rules, while adult use remains Schedule I federally. Multi site trials, the kind that produce definitive answers, have to navigate all of it.
Two studies recruiting now
A study of cannabis effects on sleep, circadian rhythms, and light sensitivity enrolls adults who use cannabis and measures, under supervision, what the products they already use do to sleep, one of the most common reasons people say they use it and one of the least studied. A trial of cannabidiol in older adults who use cannabis tests CBD for sleep, anxiety, depression, and pain in the fastest growing group of new cannabis users, adults over 60. Both are listed with eligibility and locations on our cannabis use disorder clinical trials page. Studies that provide cannabis or CBD do so under federal licensing, in controlled settings, and assign groups by randomization; they are research, not a source of medical cannabis, and every detail is explained during informed consent.
Common questions
Is marijuana rescheduled now? Partly. Since April 2026, FDA approved cannabis products and cannabis sold under state medical marijuana licenses are in Schedule III. Cannabis sold for adult use remains in Schedule I while a DEA hearing that began in June 2026 considers moving all cannabis to Schedule III. Rescheduling is not legalization: cannabis remains federally controlled either way.
What does Schedule 3 mean for cannabis? It means the federal government recognizes accepted medical use and a moderate potential for dependence, the same tier as ketamine and some codeine products. Practically, it removes the special registration and storage requirements that made research so slow, changes how medical cannabis businesses are taxed, and makes FDA drug trials far easier to run. It does not make interstate sales legal or override state law.
Is medical marijuana FDA approved? Not as a plant. The FDA has approved three cannabinoid medicines: dronabinol and nabilone, synthetic forms of THC, and Epidiolex, a purified CBD extract for two rare epilepsy syndromes. State medical marijuana programs operate outside FDA approval, which is exactly why so much of what is sold has not been tested to the standard other medicines meet.
What conditions does cannabis have real evidence for? The 2017 National Academies review found conclusive or substantial evidence for three: chronic pain in adults, chemotherapy induced nausea and vomiting, and patient reported spasticity in multiple sclerosis. Evidence for anxiety, PTSD, sleep, and most other uses was limited or insufficient, largely because the studies could not be run. Many of the trials recruiting now are filling those gaps.
Do you get cannabis through a clinical trial? Some studies provide cannabis, THC, or CBD to participants under medical supervision, either to measure its effects in people who already use it or to test CBD for a specific condition. These studies have strict eligibility rules, use federally licensed supply, assign groups by randomization, and are conducted in controlled settings. They are research, not a way to obtain medical cannabis, and each listing explains exactly what is provided and to whom.