Spotlight: cannabis use disorder treatment in 2026 — what the evidence actually says

Cannabis is in the strange position of being the most-used illicit substance in the United States, the second-most-publicly-debated substance after fentanyl, and the least-treated substance in proportion to its disease burden. Roughly one in three regular adult users meets diagnostic criteria for cannabis use disorder. As of May 2026, zero medications are FDA-approved for it. The April federal rescheduling action — moving FDA-approved and state-medical cannabis to Schedule III — does not change that fact. What it does change is the population of people who will, over the next 18 months, ask the question: Am I using too much, and if I want to stop, what actually works?

Where the treatment evidence is — psychology first

The 2026 network meta-analysis published earlier this spring pooled 57 RCTs — 21 psychosocial trials with 3,157 participants, 36 pharmacotherapy trials with 3,120 participants — in CUD across the lifespan. The clearest signal: DBT/ACT (dialectical behavioral therapy / acceptance and commitment therapy) and MET-CBT combined with contingency management were the most consistent reducers of cannabis use frequency. These are the same therapies that lead in stimulant and alcohol use disorder treatment for the same reason: they treat the behavior, the affect that drives the behavior, and the reinforcement schedule that maintains the behavior simultaneously.

Pharmacotherapy is where the field is honest about its limits. Several agents — N-acetylcysteine, gabapentin, dronabinol — have shown signal in individual trials. The NMA authors classify all of them as experimental at present and caution against routine prescribing outside research settings. The mechanism work is moving forward (the JCI 2024 review is the best recent synthesis of CB1-receptor neurobiology in CUD), but the translation from receptor pharmacology to FDA filing is years away.

Where the field genuinely disagrees

The first disagreement is diagnostic. Critics — including some addiction-medicine physicians who otherwise prescribe medications for opioid use disorder daily — argue that CUD is over-diagnosed in users who would self-correct with a brief intervention, and that the DSM-5 criteria don’t distinguish between problematic use and tolerance among medical patients. Defenders, including most of the network meta-analysis authors, argue the opposite: CUD is under-diagnosed because of cultural normalization and clinician reluctance to label cannabis as a substance use disorder at all. Both groups agree on one thing — that we lack high-quality outcomes data segmented by use intensity, route (smoked vs. edible vs. vaped), and product potency.

The second disagreement is about co-occurring mental illness. A recent 2026 analysis reports that 31% of people with CUD also meet criteria for major depressive disorder. Researchers split on directionality — whether cannabis use is contributing to depression, depression is driving cannabis use as self-medication, or both share underlying neurobiology. Clinically, the implication is the same: a CUD treatment plan that does not screen and treat for co-occurring depression is incomplete.

The third disagreement is generational. Among adolescents, Monitoring the Future trend data shows past-12-month overall cannabis use has held roughly flat — but the proportion who vape it has climbed substantially: from 47% to 63% in 8th grade, 41% to 53% in 10th grade, between 2021 and 2024. Flavored cannabis vape solutions, in particular, are the only drug-use indicator moving up in 8th grade across all 30 MTF-tracked substances. Whether that represents a public-health risk equivalent to the 2018 nicotine-vape wave is the open question the next two years of data will answer.

Why this matters for people in recovery

If you are wondering whether cannabis is taking up more space in your life than you want it to: the evidence says psychotherapy works — specifically MET-CBT, contingency management, and DBT/ACT — and the absence of an FDA-approved pill is not a reason to wait. If you are a parent watching your teenager vape with increasing frequency and decreasing curiosity about why, the same answer applies: behavioral interventions, paired with a relationship that asks the right questions, are the intervention. The April rescheduling order will, over time, free up research dollars for CUD pharmacotherapy. Today, the treatments that work are talk-therapy treatments, and they are reimbursed by AHCCCS and most commercial insurance in Arizona.

Today, the treatments that work are talk-therapy treatments, and they are reimbursed by AHCCCS and most commercial insurance in Arizona.

For treatment-matching, start with Rize. For research and policy in this space, browse our cannabinoid briefings.

If you are in crisis: 988 for the 988 Suicide & Crisis Lifeline, 911 for emergency services, or SAMHSA 1-800-662-HELP (4357) for free confidential substance-use referral 24/7.

— The Rize Recovery Newsroom