Cannabis Has Been Legal in Thirty-Eight States. There Is Still No FDA-Approved Treatment for Cannabis Use Disorder.

The hearing starts June 29 in Arlington, Virginia. The Drug Enforcement Administration will spend up to three weeks taking testimony on whether marijuana in its entirety — not just the FDA-approved products and state-licensed medical programs that moved to Schedule III in April — belongs on Schedule I of the Controlled Substances Act.

Whatever the DEA decides, it will land in a particular context: there are approximately 19.2 million Americans who met criteria for cannabis use disorder in the past year, according to NIH data updated in March 2026. The number of people developing CUD has grown by 32 percent since 2019. And as of today, June 4, 2026, there is not a single medication that the Food and Drug Administration has approved to treat it.

Cannabis has been normalized, legalized, commercialized, and rescheduled. The disorder that regular cannabis use can cause has been largely ignored.

What cannabis use disorder actually is — and why it’s harder to treat than people think

The popular image of cannabis dependence is dismissive: a pothead who needs to smoke a joint to get through the day. The clinical picture is more specific and more serious than that. The DSM-5 describes cannabis use disorder as a pattern of use leading to significant impairment, characterized by criteria that include tolerance (needing more to get the same effect), withdrawal (irritability, anxiety, sleep disruption, appetite changes, restlessness), craving, and continued use despite knowing it’s causing problems. The disorder ranges from mild to severe. At the severe end, it looks more like opioid dependence than it does like the caricature.

What the imaging research shows is equally clarifying. A 2024 review published in the Journal of Clinical Investigation synthesized the existing neurobiological literature: regular cannabis use is associated with measurable changes in the brain regions governing reward, craving, and cognitive control — the same regions implicated in every other substance use disorder. The neural response to cannabis cues in people with CUD overlaps significantly with the cue-reactivity patterns seen in opioid use disorder and alcohol use disorder. The craving is the same kind of craving. The circuitry is the same circuitry.

Approximately 60 percent of people with CUD experience significant craving, according to the same review. About 32 percent go through withdrawal. And the relapse rate — around 67 percent — is roughly comparable to relapse rates across other substance use disorders.

None of this is intuitive for a substance that a plurality of Americans now use recreationally and that many states have made easier to obtain than alcohol. The normalization of cannabis has had a useful effect on some policy conversations and a harmful effect on others. One of the harmful effects: it has made it harder for people with CUD to access treatment or even have their disorder taken seriously.

“I tried to tell my primary care doctor I thought I had a problem with weed,” one person in a recovery community forum wrote in a thread aggregated by a harm reduction researcher in 2025. “He laughed. He said, ‘That’s not a real thing.’” The pattern in these threads is consistent: people experiencing the specific constellation of cannabis withdrawal — the anxiety, the sleeplessness, the inability to feel pleasure without the drug — and being told by the healthcare system that what they’re describing isn’t real.

The research gap that rescheduling could unlock

Here is what Schedule I has meant for cannabis research: a drug classified as having no accepted medical use and a high potential for abuse requires DEA-licensed facilities, NIDA-approved supply chains, and layers of federal oversight to study. The result is that the United States has a decades-long gap in clinical-quality pharmacology research on cannabis and cannabinoid receptors. Other countries — particularly the Netherlands, Spain, Canada — have produced much of the foundational work that American researchers have had to import.

Other countries — particularly the Netherlands, Spain, Canada — have produced much of the foundational work that American researchers have had to import.

The DEA final order issued April 28, moving FDA-approved cannabis products and state-licensed medical marijuana to Schedule III, is the most significant administrative shift in decades. Schedule III means lower barriers for research, less regulatory overhead for clinical trials, and — critically — the ability for DEA-registered pharmaceutical manufacturers to study cannabinoid-based pharmacology without the extraordinary compliance architecture that Schedule I requires.

If the June 29 hearing results in full rescheduling for all marijuana to Schedule III, the research infrastructure that has been structurally blocked since the Controlled Substances Act of 1970 becomes substantially more accessible. That is not a guarantee of cures. It is a removal of obstacles that should have been removed long ago.

What is being tested now, and what’s missing

In the absence of an approved medication, clinicians treating CUD are working with incomplete tools. The evidence base for behavioral approaches — specifically cognitive behavioral therapy, contingency management, and motivational enhancement therapy — is reasonably solid. People can and do recover from CUD through these approaches. But behavioral therapy has higher dropout rates and more variable outcomes when there is no pharmacological support alongside it. The combination, in virtually every other SUD context, outperforms either approach alone.

The most promising pharmacological candidate is AEF0117, developed by Aelis Farma in collaboration with Columbia University’s research team. AEF0117 is the first compound in a new pharmacological class called CB1-SSi — selective CB1 signaling inhibitors. Rather than blocking the CB1 receptor entirely (which would be blunt-force pharmacology with significant side effects), CB1-SSi drugs target only the specific cellular signaling cascades involved in cannabis reward and dependence. The theory is that you can interrupt the addiction circuitry without disrupting the receptor’s other physiological functions.

The Phase 2b trial currently underway at Columbia will enroll 330 participants across multiple sites and test three dose levels of AEF0117 against placebo. It is a rigorous study. The results will matter enormously for whether there is a viable medication path forward. But the study will take years to complete and analyze. And even after a Phase 2b result, a Phase 3 trial, FDA review, and approval process stretches the timeline further. The people who need help today are not waiting for 2030.

Other candidates have had inconsistent results: gabapentin, N-acetylcysteine (with some signal in adolescents), naltrexone, varenicline, and CBD at high doses have all been studied with mixed outcomes. None has accumulated sufficient evidence for approval. What the field is missing is not effort — it is the research infrastructure that Schedule I has suppressed.

What it means for people looking for treatment now

For someone with CUD looking for help today, the clinical pathway is narrower than it should be. Behavioral therapy is the standard of care. Residential or intensive outpatient programs can be appropriate for severe presentations. Medications for comorbid conditions — anxiety, depression, insomnia — can be addressed pharmacologically even when the CUD itself cannot.

What matters most for outcomes, the research is consistent on this, is continued engagement with treatment. The 67 percent relapse rate is not a death sentence. It is a description of how the disorder works — and an argument for sustained, supported recovery that does not treat one course of therapy as the finish line.

The /newsroom/substances/cannabinoids section of the Rize newsroom tracks research on cannabis use disorder, treatment outcomes, and harm reduction for people using cannabis. The /newsroom/category/science-medicine section covers pharmacological developments across all substance classes.

The June 29 DEA hearing is not a cure. It is, at best, the beginning of a decade of research that should have started in 1980. What people who need help now deserve is the acknowledgment that what they are experiencing is real, that treatment exists, and that the absence of an FDA-approved medication for their specific disorder is a failure of policy and research infrastructure — not evidence that the disorder doesn’t warrant one.

It is, at best, the beginning of a decade of research that should have started in 1980.