CBD, CBN, THC and Sleep

The question I get asked more than almost any other is some version of this: What should I take to sleep better?

For years, the cannabis industry's answer has been simple. THC makes people sleepy, so THC must be the sleep cannabinoid.

I understand why that idea caught on. THC can make people drowsy, and many people who use it at night say they fall asleep more easily. Dispensaries have recommended THC-heavy products for nighttime for years, and marketing has followed the same logic.

The problem is that feeling sedated and sleeping well aren't the same thing.

That difference has become more important as better studies have started looking at individual cannabinoids instead of treating cannabis as one substance. CBD, CBN and THC are chemically related, but they don't act the same way in the body, and we shouldn't expect them to affect sleep the same way either.

Two recent research reviews show how young this science still is. They reached somewhat different conclusions, but together they point to a more useful question than whether "cannabis helps sleep." The better question is: which cannabinoid, at what dose, in what combination, and for which part of the sleep problem? That's where the evidence is starting to get useful.

A new review questioned the THC assumption

In 2026, researchers published a review in the Journal of the American Association of Nurse Practitioners that pooled studies of medical cannabis for insomnia in adults. It covered CBD, CBN, THC and combinations, taken as capsules, oils or drops under the tongue. The researchers looked at whether people said they slept better, and also at sleep disruptions, total sleep time, daytime sleepiness and the doses used.

Their conclusion didn't match the usual dispensary advice. Across the studies, CBD doses of 50 to 300 mg and CBN doses of 20 to 100 mg were linked to the most consistent improvements. CBD below 50 mg generally seemed helpful only when CBN was also present. Products containing THC didn't show the same consistent improvement and came with more side effects.

That's worth noticing, but I wouldn't turn it into a rule. A review like this combines studies that can differ a lot in dose, length, product, the people studied and how sleep was measured. A dose range that keeps showing up doesn't mean everyone should take a dose inside it, and it certainly doesn't mean more is better. What it does tell us is that the old idea of THC as the sleep cannabinoid is getting much harder to defend as a blanket statement.

Then another review complicated things

A separate 2025 review in Sleep Medicine Reviews pooled six randomized controlled trials with 1,077 people in total. Overall, cannabinoids were linked to better self-reported sleep quality than placebo, especially in people who already had insomnia or poor sleep. But the benefit came from formulas containing THC and/or CBN. CBD on its own didn't produce a statistically significant improvement.

The results also varied a lot from trial to trial. The authors cautioned that differences in products, doses and study groups limit how far the findings can be applied.

At first glance, the two reviews seem to disagree. I don't think they really do. I think this is what young science usually looks like. Different reviews use different rules for which studies to include, measure different outcomes, and group products differently. One might count CBD-plus-CBN separately; another might lump everything that isn't pure CBD together. One focuses on people with insomnia; another includes people who just sleep poorly. The results shift as the question shifts.

What matters is where the findings overlap, and one of the most interesting overlaps is CBN.

CBN finally has human data behind it

CBN, or cannabinol, has been sold as a sleep cannabinoid for years with surprisingly little human evidence behind it. That's starting to change.

A 2024 randomized, double-blind, placebo-controlled study enrolled about 300 adults who rated their own sleep as poor or very poor. Each person was randomly assigned to a placebo, 20 mg of CBN alone, or 20 mg of CBN combined with 10 to 100 mg of CBD. After seven nights, here's what 20 mg of CBN did compared with placebo:

This is useful because it pulls apart questions people usually lump together. Falling asleep is one problem. Staying asleep is another. Waking up over and over is another. Feeling foggy the next morning is another. A compound doesn't have to fix all four to help.

In this study, CBN's strongest signal wasn't that it knocked people out faster. It was that they woke up less often and reported less disruption overall. That's very different from sedation.

Sedation isn't the same as restful sleep

This may be the most important idea in the whole conversation. We tend to judge a sleep aid by how strongly we feel it before bed. If something makes us drowsy within half an hour, we assume it's working.

Sleep researchers aren't as easily impressed. They can ask what happened during the rest of the night. How long did the person actually sleep? How often did they wake up? How long were they awake after first dozing off? Did the stages of sleep change? Did they feel alert or wiped out the next day? Those don't always move together.

Alcohol is the obvious example. It can help you fall asleep faster, especially at higher amounts, but it disrupts sleep later in the night. Research has shown again and again that alcohol makes the first part of the night feel easier while making the rest of it worse. That's sedation without better sleep.

Cannabis raises a similar question. A 2025 review of cannabis and the structure of sleep found that it didn't consistently improve how long people slept, how fast they fell asleep, how long they lay awake, how efficiently they slept, or their sleep stages. Earlier studies that found big drops in REM (dreaming) sleep were often small and used fairly high THC doses; newer studies have been mixed.

That doesn't mean nobody sleeps better after THC. It means "I feel sleepy after THC" and "THC improves sleep across a group of people" are two different claims. We need to stop treating them as the same.

CBD is more complicated than its reputation

CBD has almost the opposite problem. It's often sold as a calming sleep ingredient, even when the amount in the product looks nothing like the amounts used in research. And the human data on CBD alone are mixed.

A small randomized trial in the Journal of Clinical Sleep Medicine gave people with moderate-to-severe insomnia 150 mg of CBD every night for two weeks. Compared with placebo, CBD didn't significantly improve the main sleep measurements, either those taken by instruments or those people reported themselves, although some measures of general well-being improved. A follow-up analysis found that the same nightly dose didn't impair thinking the next day. Attention, planning, reasoning and memory held steady, and CBD looked safe in the short term.

That pushes us away from the simple question of whether CBD "works for sleep." CBD may affect anxiety, a racing mind, discomfort or other things that get in the way of sleep, without acting like a classic sleeping pill. For someone who keeps waking during the night, that may be a different problem from someone whose mind races for an hour before they drift off. We shouldn't assume one cannabinoid will fix both.

What the dose research does, and doesn't, tell us

The 2026 review linked CBD doses of 50 to 300 mg and CBN doses of 20 to 100 mg to better sleep, and found that lower CBD doses tended to be more useful alongside CBN.

Those numbers matter because they expose a real problem in the supplement market. A product can contain a cannabinoid and still have far less of it than the research being used to sell it. If a company cites a study that used 100 mg and sells a product with 3 mg, you deserve to know those aren't remotely comparable.

At Blue Sky, our Sleep Oil contains 40 mg of CBD and 40 mg of CBN per dropper. Our Sleep Gels contain 25 mg of CBD and 25 mg of CBN per gel. Here's how those compare with the ranges in the review:

Those numbers help explain how the formula was put together. They don't tell you what your own dose should be. And taking two servings doesn't automatically double the benefit. The studies in these reviews didn't use our products, and cannabinoid doses don't work in a simple "more is better" way. The numbers should help you understand what was studied, not turn a research review into a prescription.

The CBN trial makes the dose question even more interesting

This is another reason I wouldn't treat the 50-to-300 mg CBD range as a target everyone needs to hit. In the 2024 CBN trial, 20 mg of CBN alone reduced nighttime awakenings and overall sleep disturbance. Adding 10, 20 or even 100 mg of CBD didn't make it work any better.

That's a good reminder to judge a formula by results, not by the biggest milligram number on the bottle. More CBD doesn't automatically make a better sleep formula. More total cannabinoids doesn't automatically make it stronger in a way that matters. What counts is what each ingredient does and whether the combination has actually been studied. We're still learning that.

Before you change a supplement, look at the rest of your day

The cannabinoid conversation is interesting, but it can distract from a basic truth: any supplement is working on top of a sleep system shaped by dozens of other signals.

I've never seen a supplement make an erratic sleep schedule stop mattering. I've never seen one cancel out heavy afternoon caffeine. And I've never seen one fix untreated sleep apnea. A formula can support sleep. It can't replace the biology that produces it. Here's what I'd look at first:

How I think about our sleep formulas

Once those basics are in decent shape, choosing a formula becomes a much more useful conversation. I don't think everyone who sleeps poorly needs the same product, because I don't think everyone who sleeps poorly has the same problem.

If waking up during the night is the main issue, I'm most interested in formulas that contain CBN, because that's where some of the most direct human evidence is appearing. Our Sleep Oil combines 40 mg of CBD with 40 mg of CBN per dropper. Sleep Gels provide 25 mg of each per gel. The oil gives you a flexible liquid format; the gels give you a fixed dose if you'd rather take a capsule than use an oil.

Neither one has been tested directly in the trials discussed here. That matters. The studies explain why we're interested in CBD and CBN and why they're in the formula. They don't let us claim our products will reproduce the study results.

If physical discomfort is what keeps you up, I'd address that rather than just adding more nighttime cannabinoids. In our line, that might mean a relief-focused formula instead of assuming every bad night needs a dedicated sleep product. The product should fit the problem. That sounds obvious, but mismatching them is one of the most common mistakes I see.

An important caveat about CBN and drug testing

If you're drug tested for work or anything else, you should know this about CBN before using it.

A 2020 lab study tested four cannabinoids against two common urine screening tests designed to detect THC's main by-product. CBD, CBC and CBG didn't meaningfully set them off. CBN did.

This finding needs to be described accurately. The researchers added known amounts of CBN directly to urine samples. They did not give people CBN sleep products and count how many failed a workplace drug test. So I wouldn't tell anyone that taking CBN will cause a positive result; the evidence doesn't show that.

What it does show is that CBN can set off at least some common screening tests. If a screening result doesn't match what you expect, a confirmation test (mass spectrometry) can tell CBN and THC's by-product apart, because they're different molecules. But if passing a drug test matters for your job, take that uncertainty seriously. We can't promise you a result on somebody else's test.

What "0.0% THC" means

Our approach at Blue Sky is to remove THC and have finished products independently tested. Our products are labeled 0.0% THC, and each batch has a lab report, called a certificate of analysis, that backs up that number.

It's worth being precise about what that means. 0.0% THC on our label reflects that the lab didn't find THC at the level its test can measure. It isn't proof that not a single molecule exists. No lab instrument can prove absolute zero.

That can sound overly technical until drug testing or workplace rules matter to you. Then it becomes very practical. Read the certificate for the batch you actually have, understand what the lab measured, and be skeptical of any company that claims its testing proves more than the test can show.

Judge the night by the morning

There's one practical idea I'd keep from all this research even if cannabinoids disappeared tomorrow: don't judge your sleep only by how fast you fell asleep. Ask how the night went:

One reason I'm interested in the CBN trial is that the researchers measured next-day tiredness along with the nighttime results. CBN reduced awakenings and sleep disturbance without making people more tired the next day. That's the kind of sleep result I care about.

If something knocks you out faster but leaves you foggy for half the next day, I'm not convinced it improved your sleep. Sleep exists partly to set up the next day. The morning is where you find out whether the night actually worked.

Where the research leaves us

I don't think the evidence lets us crown a winner among CBD, CBN and THC. The science isn't mature enough for that. What it does support is dropping the habit of treating all cannabinoids as interchangeable.

The 2026 review found the most consistent results with CBN, and with CBD at certain doses or alongside CBN, while THC formulas were less convincing and had more side effects. The 2025 review found that cannabinoids helped overall but CBD alone didn't, and its studies varied enough that nobody should boil the research down to a slogan. The CBN trial gave one of the clearest signals so far: 20 mg of CBN reduced nighttime awakenings and sleep disturbance, and adding CBD didn't improve those results. Meanwhile, controlled studies of CBD alone remain mixed, and the measured research on cannabis and sleep structure is far less decisive than decades of personal stories would suggest.

That's not a disappointing conclusion. It's progress. For years, this category ran on a simple story: THC makes you sleepy, CBD makes you calm, and CBN is somehow the sleep cannabinoid. Now researchers are starting to ask what each molecule actually does.

That's the conversation I want to have. Not which cannabinoid makes you feel the most sedated, but which one helps produce a better night, for the right person, without making the next day worse. We don't have the full answer yet. But at least we're finally asking the right question.

Eric Dorninger, ND
Director of Research & Development, Blue Sky CBD