Does Cannabis Help Insomnia? What Nine Trials Show

Can medical cannabis help people with insomnia sleep better and function better the next day? That question sounds straightforward, but answering it requires more than showing that a product makes someone feel sleepy. A useful treatment must produce meaningful improvements, with benefits that justify its unwanted effects.

A new Cochrane review, “Medical cannabis for the treatment of insomnia,” examines that question across nine randomized trials involving 537 adults.¹ Published in October 2026, it finds that the evidence remains too uncertain to establish reliable benefits for sleep quality, sleep duration, or daytime functioning.

That conclusion provides a useful direction for research. Rather than treating cannabis as one universal sleep aid, scientists need to identify which formulations might help particular sleep problems, for whom, and for how long.

Why Insomnia Requires More Than Feeling Sleepy

Insomnia can involve difficulty falling asleep, staying asleep, or waking too early, together with problems during the day. Chronic insomnia involves recurring difficulties over at least three months. The consequences extend beyond an unpleasant night to concentration, energy, mood, and everyday functioning.

These distinctions matter when evaluating cannabinoids. Tetrahydrocannabinol, or THC, and cannabidiol, or CBD, are different compounds. Products containing them vary in composition, dose, and delivery method. Evidence about one preparation cannot automatically establish the effects of another.

Sleep itself also has several dimensions. Someone might fall asleep faster without sleeping longer. Another person might report a more satisfying night while still struggling with daytime tiredness. Research needs to examine these outcomes separately before combining them into a broad claim that something improves sleep.

How the Cochrane Cannabis Review Worked

The researchers searched medical databases and trial registers through April 5, 2025. Although the review was published in 2026, its search cutoff matters: the publication date does not mean it captures every finding released during that year.

The nine included trials enrolled adults aged 18 to 76. Eight compared medical cannabis with placebo, an inactive treatment. One compared it with the antidepressant amitriptyline. Six studies involved chronic insomnia and three involved acute insomnia. Most participants had primary insomnia, rather than insomnia associated with another medical condition.

Seven trials used crossover designs, meaning participants received different treatments during separate periods. This allows researchers to compare a person’s response under different conditions, but it also requires careful handling of treatment order and possible lingering effects.

The review assessed both results and their reliability. Small samples, inconsistent findings, and problems that could bias results lowered confidence in the evidence. Combining studies can improve understanding, but it cannot erase weaknesses in the underlying experiments.

What the Review Found About Cannabis and Sleep

Compared with placebo, medical cannabis showed no clear, clinically meaningful improvement in sleep quality, total sleep time, or daytime functioning. The researchers rated the certainty of these findings as very low.

For total sleep time, the pooled estimate was approximately eight additional minutes. However, the uncertainty interval stretched from about 23 fewer minutes to 38 additional minutes. That range includes substantially different possibilities, making the average alone a poor basis for a confident promise.

Quality of life showed a possible improvement, but this evidence was also very uncertain. Adverse events were more frequent with cannabis, with low-certainty evidence supporting that finding.

The differing participant totals reflect which studies contributed usable data to each analysis. They should not be added together as separate groups of patients.

What Uncertain Evidence Actually Means

There is an important difference between failing to establish a benefit and establishing that no benefit exists. Here, the studies leave too much uncertainty for either a broad endorsement or a definitive rejection.

In its explanation of the insomnia findings, Cochrane emphasizes that the available research cannot reliably settle effectiveness. That leaves room for better trials while putting limits on claims made today.

Think of the evidence as a photograph taken in poor light. A blurry image does not prove that nothing is there. Equally, it cannot justify describing details that remain invisible. Larger, better-controlled studies should sharpen the picture.

Personal experiences still deserve attention. They can identify outcomes that matter to patients and suggest questions worth testing. However, an individual improvement cannot reveal how much came from the product, expectations, changing symptoms, or other changes in routine.

Why Cannabis Sleep Research Needs More Precise Questions

A useful next step is to replace “Does cannabis work for sleep?” with narrower questions. Does a particular formulation reduce nighttime waking? Does it help people whose sleep is disrupted by pain? Does improvement persist after several months?

A 2026 placebo-controlled CBN insomnia trial illustrates why those distinctions matter. Twenty adults received cannabinol, or CBN, at 30 milligrams, 300 milligrams, and placebo on separate treatment nights. Neither dose significantly improved the primary outcome, time spent awake after initially falling asleep.

However, the higher dose improved some secondary outcomes, including time to fall asleep and reported sleep quality. Those findings support further investigation, but do not establish an effective routine dose. When the main outcome is negative and other measures improve, replication becomes especially valuable.

This is also why consumer satisfaction and controlled efficacy should remain separate. MyCannabis’s recent coverage of CBN use and the dosing evidence gap examines experiences that can inform research questions. Such observations help describe what people are doing; randomized comparisons test whether a treatment itself produces an improvement.

Long-Term Benefits Need Long-Term Testing

Insomnia can persist far longer than the trials designed to study it. The Cochrane review identifies short treatment periods and limited reporting of withdrawal and other longer-term outcomes as important gaps.

A treatment could appear helpful initially yet become less useful with repeated exposure. Alternatively, a sustained benefit might emerge only in a particular patient group. Short studies cannot confidently distinguish those possibilities.

Recent MyCannabis coverage of CBN tolerance and dependence research highlights why repeated use deserves its own investigation. Animal findings cannot establish the risk from ordinary human sleep products, but they help identify questions human trials should address.

Future studies would be more useful if they consistently answered three questions:

• Does the exact formulation improve a clearly defined sleep problem?
• Do benefits persist alongside acceptable daytime functioning?
• What happens during prolonged use and after treatment ends?

What the Findings Mean for CBD Sleep Products

The practical implication is to keep product descriptions separate from clinical evidence. A “sleep” label does not establish that a formulation treats insomnia, and results from a tested preparation cannot simply be transferred to every oil or gummy containing similar ingredients.

For readers comparing labels, the MyCannabis CBD sleep products buying guide explains differences among formulations, including products containing additional ingredients such as melatonin. It is a product-literacy resource, rather than evidence that a listed product treats insomnia.

This distinction creates a constructive standard for manufacturers: test the actual product, specify the intended outcome, and report benefits and unwanted effects together. For persistent sleep difficulties, clinical assessment remains more informative than relying on a marketing category alone.

The Next Step for Cannabis Insomnia Research

The review gives cannabinoid research a clearer agenda. Future trials need larger and more diverse groups, consistent measurements, well-characterized formulations, and enough follow-up to assess sustained benefit. Older adults and people with additional health conditions also need better representation.

The most useful next experiment would test a defined preparation against placebo over a meaningful period, measuring both nighttime sleep and the following day. It should also examine what happens after treatment stops. That would move the conversation toward the question patients actually need answered: whether a particular treatment makes their lives measurably better.

References:

¹ Varela, L. B., Miño Zambrano, J. E., Maura Goncalves Costa, B., Santos Marques da Silva, G., Escobar Liquitay, C. M., Díaz Menai, S., Meza, N., & Garegnani, L. I. (2026). Medical cannabis for the treatment of insomnia. Cochrane Database of Systematic Reviews, 2026(10), Article CD016216. https://doi.org/10.1002/14651858.CD016216.pub2