Walk the sleep shelf at any dispensary or hemp shop and the verdict is already printed on the tins. CBN is the sleep cannabinoid, CBD is the calm one, and the premium SKU is the one that combines them. The trial literature, which only started to exist in 2024, agrees with about half of that.
Here is the half it agrees with, and our verdict. If you are choosing between a CBN gummy and a CBD gummy to sleep, the evidence favors CBN, at 20mg and up, and only for staying asleep rather than falling asleep. CBD at the 25mg printed on most sleep gummies has no controlled trial behind it: the sedation evidence for CBD starts at 160mg, and a 2024 trial at 150mg nightly found it no different from placebo on insomnia severity. CBD earns a place on the nightstand only when anxiety is what keeps you awake. And the combination gummy, the shelf's premium tier, has now been tested from both directions. Adding CBD to CBN did nothing (Bonn-Miller 2024), and adding CBN to CBD did nothing (Saleska 2024). Paying extra for the combo is paying for the one formulation two industry-funded trials agree is inert.
One number frames the piece. The lowest CBN dose that has ever beaten a placebo is 20mg. Kiva's Camino Midnight Blueberry, which Kiva's own page calls its top-selling product, carries 1mg of CBN beside 5mg of THC, chamomile and lavender. Our CBN gummies ranking grades the shelf against the 20mg floor; this page is the evidence underneath it.
Does CBN actually help you sleep?
Yes, for staying asleep, on thin evidence. The only placebo-controlled CBN result at a dose a gummy can reach is Bonn-Miller 2024, where 20mg nightly cut nighttime awakenings and overall sleep disturbance, missed its primary endpoint and did nothing for time to fall asleep. Kolobaric 2024 found 25 to 100mg no better than 4mg of melatonin.
Bonn-Miller randomized 293 adults who rated their own sleep as poor or very poor to placebo, 20mg CBN, or 20mg CBN with 10, 20 or 100mg of CBD, taken as two gummies 90 minutes before bed for seven nights. The CBN-only arm reduced the number of awakenings (p=.025) and overall sleep disturbance (p=.023), fell short on the primary sleep-quality score (p=.082), and showed no difference from placebo on sleep onset latency, minutes awake after first waking, or daytime fatigue. The authors came from Canopy Growth and Wana Brands, and Wana sells CBN sleep gummies. Hold that beside a nonsignificant primary endpoint and the result reads as what it is: a real but narrow effect on how often you wake, with nothing on how fast you fall asleep. Our sleep maintenance page builds its protocol on that one arm.
Kolobaric 2024 is the large one: 1,020 adults, four weeks of nightly softgels in a fully remote trial, with 25, 50 and 100mg of a hemp-derived CBN called TruCBN against 4mg of melatonin and placebo. Every active arm beat placebo on the PROMIS sleep disturbance score. No CBN dose separated from the melatonin, which the paper says twice, and which is the sentence to remember when a $40 jar of CBN gummies sits next to a $3 drugstore bottle. Floraworks, which makes TruCBN, paid for the study, and Radicle Science employees ran and wrote it.
Lavender 2026 is the trial with a sleep lab. Twenty adults with physician-diagnosed insomnia disorder spent three separate nights at the Woolcock Institute in Sydney on 30mg of CBN oil, 300mg, or placebo, wired for polysomnography. The primary outcome, minutes awake after sleep onset, did not move at either dose. The 300mg dose cut sleep onset latency by about seven minutes (the paper calls it halving), added N2 sleep and improved subjective sleep quality. The 30mg dose, the realistic end of the hemp shelf, did nothing. The 300mg dose also produced a small, measurable drug effect, with no change on the "feel high" scale, that the authors put at roughly a tenth of what THC produces. The Lambert Initiative, a philanthropically funded cannabinoid program at the University of Sydney, paid for it.
That is the entire placebo-controlled human literature on CBN and sleep: three trials, two of them paid for by companies that sell CBN, and one sleep lab with 20 people in it. Corroon's 2021 review, written before any of them, found nothing in the record and noted that products marketed for sleep typically carry 5mg or less.
Why would eating CBN work better than the receptor math says?
Because the liver turns CBN into something closer to THC. Arnold 2025 found that in rats, the metabolite 11-hydroxy-CBN reached brain concentrations equal to CBN itself and activated the CB1 receptor with potency and efficacy comparable to THC, where CBN alone barely registers. Swallowing a dose sends all of it through that conversion first.
The Sydney group published this in Neuropsychopharmacology in February 2025, and it is the most interesting CBN paper of the decade for anyone who sells gummies. In rats on wireless polysomnography, 10 mg/kg of CBN increased non-REM sleep by about as much as 10 mg/kg of zolpidem, with two differences: the CBN effect arrived three to four hours after dosing where zolpidem's was immediate, and CBN did not shorten the time to fall asleep. At the CB1 receptor, CBN managed a maximal response of 29 percent; its metabolite managed 75 percent, a partial agonist a little less potent than THC, and dosed on its own it increased non-REM sleep too.
Now the edibles argument, which nobody in the consumer press has made. 11-hydroxy-CBN is made in the liver, and the oral route sends the whole dose through the liver before it reaches the brain. That is the same first-pass effect that turns swallowed THC into 11-hydroxy-THC and makes a 10mg gummy hit harder than a 10mg puff, the chemistry our drug-testing piece walks through from the metabolite side. If CBN works at all, a gummy is the format most likely to show it, and the rat timing fits what the gummy trials saw: a delayed effect in the back half of the night rather than at lights out. The caveats are the usual ones. Rats, injected, at doses no gummy approaches, in a paper whose senior authors hold cannabinoid patents, which they disclose.
The same chemistry cuts the other way for the word "non-intoxicating" on the tin. A mouse study in Drug and Alcohol Dependence this September found CBN fully substituted for THC in mice trained to tell THC from saline, and Lavender's insomniacs felt something small at 300mg. CBN is weakly THC-like, more so after your liver has had a go at it, and the label is describing a dose rather than a molecule.
Does CBD help you sleep, or does it just calm you down?
It calms you down, and the sleep benefit follows only when anxiety was the problem. No placebo-controlled trial has shown a sleep effect from CBD at the 25mg in a typical gummy. Sedation appears at 160mg in Carlini's 1981 insomnia study, and 150mg nightly was no different from placebo on insomnia severity in Narayan 2024.
The old studies set the floor. Carlini and Cunha gave 15 insomniac volunteers 40, 80 or 160mg of CBD against placebo and 5mg of nitrazepam. Only the 160mg group reported sleeping significantly more; 40 and 80mg did nothing for sleep. Zuardi's 1993 volunteers took 300 or 600mg in the morning and rated themselves sedated. Linares 2018 put 27 healthy sleepers through a night of polysomnography on 300mg and found no effect on anything. The smallest dose with a sleep signal in that literature is six gummies.
The citation every 25mg sleep gummy leans on is Shannon 2019, and it deserves to be read rather than cited. It is a retrospective chart review at a Colorado psychiatric clinic: 72 adults, nearly all on 25mg of CBD a day in capsules, as an add-on to whatever treatment they were already getting, with no control group and no blinding. Anxiety scores fell in 57 of the 72 within the first month and stayed down. Sleep scores improved in 48 within the first month and then, in the authors' own word, fluctuated. The sleep gummy industry quotes the 48 and leaves out the rest.
The modern trials are small and mostly flat. Narayan 2024 gave 30 adults with moderate to severe insomnia 150mg of sublingual CBD or placebo nightly for two weeks. Insomnia severity, self-reported sleep onset, sleep efficiency and minutes awake after sleep onset did not differ between groups; the CBD group did show better actigraphy-measured sleep efficiency after the second week, by about seven points, and higher well-being throughout, which the authors read as a psychological effect rather than a hypnotic one. Gournay 2026 gave 63 high trait worriers 300mg, 50mg or placebo for two weeks: 300mg beat 50mg on sleep disturbance and did not beat placebo. Wang 2025, funded by the company that makes the product, gave 125 insomniacs 300mg of CBD with eight terpenes and measured a 1.3 percentage-point rise in deep plus REM sleep on a wrist tracker, with no change in total sleep time. The one positive low-dose signal is Kisiolek 2023: 50mg a day for eight weeks improved a sleep-quality questionnaire score in 28 healthy college students, 14 per arm, with no change in how much they slept.
Two meta-analyses a year apart pooled overlapping trials and reached opposite verdicts. da Silva 2025 in Sleep Medicine Reviews found the sleep-quality effect sat entirely in the non-CBD arms, with CBD-only therapies showing no significant effect. Begazo 2026 in the Journal of the American Association of Nurse Practitioners credited CBD at 50 to 300mg and CBN at 20 to 100mg, with CBD under 50mg working only with CBN added. Read either one and the 25mg CBD gummy sits outside the dose range any reviewer found a signal in.
What CBD has is an anxiety mechanism and an anxiety literature. Russo's 2005 cell study showed CBD acting as a modest-affinity agonist at the 5-HT1A serotonin receptor, and Carrier 2006 showed it blocking adenosine reuptake, both at concentrations a gummy does not produce in a human brain. The anxiety trials Blessing's 2015 review summarized ran at 300 to 600mg, acute dosing, with the public-speaking paradigm doing most of the work. So the honest path from CBD to sleep runs through anxiety: a person who cannot sleep because they cannot stop worrying has a target CBD plausibly hits, at a dose far above the gummy, and the sleep comes second. Our anxiety ranking covers that use case; the THC versus CBD guide covers the rest of the molecule.
The shelf has noticed. Medterra's Sleep Tight gummy lists 25mg of CBD, 3mg of melatonin and 10mg of 5-HTP per piece at $45.99 for 30, which is a melatonin gummy with a CBD letterhead. On the same product page, Medterra now offers a melatonin-free alternative: Better Sleep TruCBN gummies, 50mg of CBN and 10mg of CBD per piece, $39.99 for 20, "clinically studied TruCBN" in the copy. A CBD company moved its sleep claim onto CBN at the exact dose from the Kolobaric trial. Nobody wrote a press release about it.
What about gummies that combine CBN and CBD?
Two trials tested the combination from opposite directions and both came back flat. Bonn-Miller 2024 added 10, 20 or 100mg of CBD to 20mg of CBN and improved nothing over CBN alone. Saleska 2024 added 15mg of CBN to 15mg of CBD in 1,793 adults and found no difference. The premium tier is a null result at a higher price.
Saleska is the Radicle Science trial that preceded Kolobaric, with two of the same authors: six capsule products over four weeks, built from 15mg of CBD isolate, 15mg of CBN, 5mg of CBC and 5mg of melatonin in different combinations, scored weekly on the PROMIS sleep disturbance scale. There was no placebo arm; every product was compared to the CBD isolate. All of them improved from baseline, which in a trial without a placebo is what everything does, and no formulation beat plain CBD. Melatonin matched it too. The shelf's logic, that CBN handles the staying asleep and CBD handles the calm, has been tested by adding each molecule to the other, and neither addition registered.
Wyld sells the cleanest version of this experiment, without meaning to. The dispensary Elderberry gummy is 10mg THC and 5mg CBN; the hemp version on wyldcbd.com swaps the THC for 25mg of CBD and keeps the 5mg of CBN, $39.95 for 20. The 1,783 Judge.me reviews on that page average 4.45 stars and include a small group who have used both. Theresa Brennan Day, a verified buyer writing on June 2, 2026, had the THC version first and found it worked great for sleep, then did not get the same effect without it. Hilliary M. ordered the hemp one by mistake on March 24 and reported it does not work like the one with THC. Andrianna H., on May 1, wrote that the THC version helps her immensely and the CBD version minimally. Brenda H., on May 13, found it takes two; one does nothing. Same brand, same fruit, same 5mg of CBN. The only variable is whether the other milligrams are THC or CBD, and the people who ran that comparison on themselves landed where the trials did.
The melatonin version of the combo deserves its own sentence. A CBN gummy with melatonin in it tells you nothing about the CBN, since 1mg of melatonin produces drowsiness on its own, and Anderson 2025 adds a twist: in mice, CBN inhibited the liver enzyme (CYP1A2) that clears melatonin and raised melatonin exposure about four-fold. On that data a CBN-plus-melatonin gummy is a melatonin gummy with a melatonin booster, a defensible product and an indefensible label.
Is old weed really more sedating?
Unproven in both directions. CBN forms when THC oxidizes, so aged cannabis carries more of it, and the folk inference was that stale flower's heavy, sleepy effect is the CBN. Russo's 2011 review offers a competing explanation: drying and storage cost the volatile monoterpenes and shift the mix toward sesquiterpenoids. No trial has compared aged flower to fresh.
Russo's Taming THC, in the British Journal of Pharmacology, holds the primary version of both stories. CBN is described as a non-enzymatic oxidative by-product of THC, more prominent in aged samples. Pages later, the monoterpenoids are described as suffering diminished yields with drying and storage, leaving a higher relative proportion of sesquiterpenoids, especially caryophyllene. The sedative terpene Russo nominates for couch-lock is myrcene, a monoterpene, which is the class storage strips out. So the terpene story does not rescue the old-weed story; it moves the question.
The human data from the era when CBN was first isolated points the other way from the lore. Corroon's 2021 review tabulated it: Hollister gave healthy men 20 to 400mg of oral CBN and observed none of the mental or physical effects of THC at any dose. In Karniol's 1975 interaction study, five volunteers felt drowsy on 25mg of THC and not on 50mg of CBN alone, though two of the THC-plus-CBN combinations rated drowsier than THC by itself, the one finding from that decade a modern gummy formulator could cite. Three trials later, "CBN does nothing" is also too strong. The marketing line Corroon quoted, that CBN's sedative properties are up to 10 times stronger than prescription sleeping drugs, had no source in 2021 and has none in 2026.
Which trial doses do gummies actually reach?
Only the hemp shelf reaches a CBN trial dose. Dispensary CBN gummies carry 1 to 10mg against a 20mg trial floor; hemp CBN gummies at 30 to 50mg clear it in one piece. No CBD gummy reaches the 150 to 300mg the trials used, and those trials were mostly null anyway. The table puts each study beside the shelf.
One caveat applies to every row: the CBN trials used softgels, gummies and an oil, and the CBD trials used capsules and sublingual drops, so the shelf equivalents are milligram arithmetic rather than a claim that a pectin gummy behaves like a trial capsule. For timing, any gummy here starts in 30 minutes to 2 hours, peaks at 2 to 3 hours and runs 4 to 8 hours for the main effects, per the figures on our edible duration guide. Bonn-Miller's participants took their CBN 90 minutes before bed; Narayan's took CBD 60 minutes before.
| Study | Dose and design | Sleep result | Shelf equivalent |
|---|---|---|---|
| Carlini and Cunha 1981 | 40, 80 or 160mg CBD against placebo and 5mg nitrazepam, 15 insomniac volunteers | 160mg: significantly more sleep reported. 40 and 80mg: no sleep effect. Less dream recall at all three doses. | Six to seven 25mg CBD gummies |
| Zuardi 1993 | 300 or 600mg CBD, 11 healthy volunteers, morning sessions | Self-rated sedation at both doses; the normal morning fall in cortisol was blunted. | 12 to 24 gummies |
| Linares 2018 | 300mg CBD, 27 healthy volunteers, one night of polysomnography | No effect on any sleep measure. | 12 gummies |
| Shannon 2019 | 25mg CBD a day (nearly all patients), 72 psychiatric outpatients, chart review, no control group | Sleep scores improved in 48 of 72 in the first month, then fluctuated. Anxiety scores fell in 57 and stayed down. | One gummy. The only study the shelf dose matches, and it had no placebo. |
| Narayan 2024 | 150mg CBD nightly for two weeks, 30 adults with moderate to severe insomnia | No difference from placebo on insomnia severity, self-reported onset, efficiency or time awake. Actigraphy efficiency better at week two; well-being higher. | Six gummies |
| Bonn-Miller 2024 | 20mg CBN, alone or with 10, 20 or 100mg CBD, seven nights, 293 self-rated poor sleepers | Fewer awakenings and less overall disturbance. Primary sleep-quality endpoint missed. No onset effect. CBD added nothing. | 20 Camino Midnight Blueberry, four Wyld Elderberry, two Camino Blackberry Dream |
| Kolobaric 2024 | 25, 50 or 100mg CBN against 4mg melatonin and placebo, four weeks, 1,020 adults | Every active arm beat placebo on sleep disturbance. No CBN dose separated from melatonin. | One Medterra TruCBN gummy (50mg) |
| Saleska 2024 | 15mg CBD alone or with 15mg CBN, against 5mg melatonin, four weeks, 1,793 adults, no placebo arm | Every arm improved from baseline. Adding CBN to CBD changed nothing. Melatonin matched CBD. | One Wyld CBD Elderberry (25mg CBD + 5mg CBN) |
| Lavender 2026 | 30 or 300mg CBN oil, single night, polysomnography, 20 adults with diagnosed insomnia disorder | Primary endpoint (minutes awake after sleep onset) null at both doses. 300mg cut sleep onset by about 7 minutes and added N2 sleep. 30mg did nothing. | 300 Midnight Blueberry gummies, or six 50mg hemp gummies |
| Suraev 2026 | 10mg THC + 200mg CBD, single dose, polysomnography, 20 adults with insomnia disorder | Total sleep down 24.5 minutes. REM sleep down 33.9 minutes. Minutes awake after sleep onset unchanged. | No gummy. Included because it is the THC result. |
The last row is the new data point, and it is about THC. Suraev 2026, from the same Sydney group, gave 20 adults with insomnia disorder a single dose of 10mg THC with 200mg CBD and measured them overnight: total sleep time fell by 24.5 minutes against placebo, REM sleep fell by 33.9 minutes, and minutes awake after sleep onset did not change. One night, one dose, 20 people, and the opposite of what "cannabis helps you sleep" promises. The THC onset evidence the site leans on is Walsh 2021, a sublingual extract delivering 10mg THC with 1mg CBN and 0.5mg CBD; both trials are on our sleep onset page, and the 1mg of CBN in Walsh's formula, the same as a Midnight Blueberry, gets credit from nobody.
So the buying rule is three lines. If the problem is waking at 3am, the molecule with a placebo-controlled result is CBN at 20mg or more, which on the licensed shelf means two Camino Blackberry Dreams (10mg CBN each, with 10mg of THC and 10mg of CBD riding along) and on the hemp shelf means one 50mg gummy while hemp CBN stays legal; our sleep ranking sorts the products by that split. If the problem is a mind that will not stop, the dose where CBD has anxiety evidence is 300mg and up, no gummy sells it, and the trials that moved sleep onset used THC. If the tin says CBN plus CBD, you are paying for the CBD, and two trials say the CBD is along for the ride. Our dosing guide covers the THC in every formula above, which remains the part of the gummy you will feel.