A large randomized clinical trial at Charité – Universitätsmedizin Berlin has found that dronabinol, a prescription form of delta-9-tetrahydrocannabinol, can substantially reduce trauma-related nightmares in adults with post-traumatic stress disorder. Published August 5, 2026, the German multicenter study enrolled 171 adults with PTSD and recurring nightmares and assigned them to receive dronabinol or placebo nightly for 10 weeks. Participants taking the cannabinoid experienced a significantly greater decline in nightmare frequency and intensity, providing some of the strongest evidence yet that a standardized THC medicine may help a specific, disabling symptom of PTSD.
More than one-third of participants assigned to dronabinol reported no remaining nightmares by the end of treatment, while another 21 percent experienced at least a 50 percent reduction in nightmare burden. About five in six said their overall health had improved. Yet the drug did not produce conclusive advantages on the principal measures of total PTSD severity or depression. This was not evidence that THC cured PTSD, nor was it a trial of dispensary cannabis, smoking, vaping, or self-directed use. It tested a standardized oral medication aimed at one symptom that often persists despite treatment.
How the Trial Was Conducted
Participants were randomly assigned under double-blind conditions to receive dronabinol or a matching placebo. The active drug was purified, plant-derived THC dissolved in oil and taken as drops roughly one hour before bedtime. Doses were adjusted between 2.5 and 15 milligrams during a four-week titration period, followed by six weeks of maintenance treatment. The primary outcome combined nightmare frequency and intensity using the nightmare item from the Clinician-Administered PTSD Scale, producing a score from zero to eight. Participants averaged about 38 years old, and nearly four-fifths were women.
Nightmare scores fell by an average of 3.7 points with dronabinol and 2.2 points with placebo. The adjusted difference was 1.50 points in favor of the drug, with a 95 percent confidence interval of 0.71 to 2.28 and a standardized effect size of 0.65. Sensitivity analyses produced similar results, reducing concern that missing data or treatment discontinuations explained the result. The sizable placebo response still matters: clinical attention, expectation, symptom fluctuation and participation in a structured trial may have contributed to improvement in both groups.
Why THC Could Affect Trauma-Related Dreams
The biological rationale begins with the endocannabinoid system, which contributes to stress regulation, fear learning, emotional memory, arousal and sleep. THC activates cannabinoid receptors, particularly CB1 receptors in the central nervous system, and may alter how threatening memories are reactivated during sleep. PTSD nightmares can recreate trauma, trigger physical arousal, fragment sleep and produce fear of going to bed. Reducing dream intensity or emotional reactivity could interrupt a cycle in which poor sleep worsens anxiety, irritability, concentration and vulnerability to intrusive memories.
However, the familiar claim that THC simply suppresses rapid eye movement, or REM, sleep is too simple. Reviews of cannabis and sleep research have found inconsistent effects on sleep architecture, with findings varying by dose, cannabinoid composition, previous exposure, duration of use and withdrawal. Acute THC may shorten sleep onset or alter REM-related activity, while continued use can reduce sleep quality and discontinuation may produce vivid dreams or REM rebound. The trial establishes an effect on reported PTSD nightmares, but it does not explain the mechanism, prove that normal sleep architecture improved or show that the benefit persists after treatment stops.
A Stronger Result Than Earlier Cannabinoid Studies
The findings build on a small and uneven research base. A preliminary 2015 randomized crossover trial found that nabilone, a synthetic cannabinoid, reduced persistent PTSD nightmares in Canadian military personnel. That study was encouraging but small and limited to men with military-related trauma. Observational reports have also described cannabis use for insomnia, hyperarousal and nightmares, but such studies are vulnerable to inconsistent products, self-selection, expectancy and confounding. The new trial carries more weight because it was larger, multicenter, placebo-controlled and organized around a clearly defined symptom.
Cannabinoid research has not otherwise demonstrated broad, reliable improvement in PTSD. In a 2021 randomized trial involving 80 U.S. veterans, smoked cannabis preparations containing different THC and CBD ratios did not outperform placebo cannabis on the primary measure of overall PTSD severity during the first treatment stage. A 2026 meta-analysis of 54 randomized trials involving 2,477 participants across mental disorders found no significant pooled benefit for PTSD and judged much of the evidence to be low certainty, while cannabinoids increased adverse events. The new study does not erase those results. It suggests that standardized cannabinoids may work better for a narrow target such as nightmares than as general treatments for every PTSD symptom.
Where Dronabinol Could Fit in PTSD Care
Trauma-focused psychotherapies remain central PTSD treatments because they address avoidance, conditioned fear, traumatic memories and distorted beliefs rather than merely suppressing symptoms. Yet sleep disruption may persist during or after therapy, and some patients cannot immediately tolerate intensive trauma processing. Imagery rehearsal therapy and other nightmare-focused approaches can help, but access and response vary. Medication options are also imperfect. Prazosin, an alpha-1 adrenergic blocker often prescribed off label for PTSD nightmares, performed well in several smaller studies but failed to beat placebo in a large 2018 trial involving 304 military veterans. Current VA and Department of Defense guidance recommends trauma-focused psychotherapy over medication for overall PTSD while suggesting prazosin specifically for nightmares based on low-quality evidence.
Dronabinol could eventually become an adjunct for selected patients whose nightmares remain severe despite established care. Its apparent specificity may be useful: better sleep could improve daily functioning and make psychotherapy easier to pursue. But the lack of clear improvement in total clinician-rated PTSD severity or depression means it should not be presented as a substitute for trauma-focused treatment. A medicine that reduces nightmares without resolving avoidance, guilt, hypervigilance, dissociation or daytime re-experiencing is addressing an important part of PTSD, not the entire disorder.
Safety Findings Deserve Close Attention
Adverse events occurred in 89.7 percent of dronabinol recipients and 77.1 percent of placebo recipients. Most were described as mild or moderate, and discontinuation because of adverse events was not higher with the drug: 5.7 percent stopped dronabinol for this reason, compared with 7.2 percent on placebo. Oral THC can produce dizziness, fatigue, appetite changes, anxiety, impairment and cardiovascular symptoms, making gradual titration and medical supervision important. This controlled protocol was fundamentally different from using an unstandardized cannabis product.
The paper also recorded seven serious adverse events in the dronabinol group and none with placebo. Investigators judged them unrelated to the medication, while peer-review documents add context about a hospitalization for tachycardia and two events coded as study-drug overdoses. One involved intentional excess intake during behavioral dysregulation, and another involved repeated modest dosing above instructions; reviewers reported no clear acute THC toxicity. These events do not prove causation, but the one-sided imbalance deserves transparent examination in larger trials and prevents the safety story from being reduced to a claim that nothing medically serious occurred.
The study lasted only 10 weeks, so it cannot establish long-term effectiveness, tolerance, dependence risk, withdrawal effects or relapse after discontinuation. Previous cannabis exposure was not systematically measured, food intake around dosing was not standardized, and sleep-disordered breathing was not comprehensively assessed. Psychoactive effects may also have allowed some participants to guess their assignment, although sensitivity analyses did not indicate that guessed treatment explained the benefit. The investigator-initiated trial received an unrestricted grant and medication from Bionorica; the authors reported that the company had no role in trial design, analysis, interpretation or publication decisions.
A Promising Symptom Treatment, Not a Verdict on Cannabis
The most defensible conclusion is hopeful but restrained. A standardized oral THC medicine produced a statistically robust and clinically meaningful reduction in PTSD nightmares in one of the largest controlled cannabinoid trials focused on this symptom. For people whose nights are repeatedly overtaken by traumatic dreams, restoring sleep may improve safety, relationships, functioning and the ability to participate in broader treatment. The study also shows why researchers must move beyond asking whether “cannabis works for PTSD” and instead test specific compounds, doses, delivery methods and clinical targets.
Longer trials should compare dronabinol with existing nightmare therapies, track cognition and daytime functioning, examine relapse after discontinuation and identify who benefits without unacceptable effects. Researchers must also determine whether lower or intermittent dosing can preserve benefit while limiting impairment and tolerance. For now, the evidence supports a narrower conclusion: medically supervised dronabinol can relieve trauma-related nightmares for some adults with PTSD, but it does not justify unsupervised cannabis use or claims that THC treats the disorder as a whole.
