Chronic insomnia is two conditions sharing one label. The first is sleep-onset insomnia, where falling asleep takes 45 minutes or longer night after night. The second is sleep-maintenance insomnia, where falling asleep is easy but staying asleep through the night is not, with awakenings at 2 or 3am that last 30 minutes or more. They share a diagnostic code (ICD-10 F51.01) and they share a name. They do not share a treatment.
This matters because most generic insomnia content treats both as one problem and recommends one solution. Take something at bedtime, stay asleep. That logic works for one of the two subtypes and fails for the other, because most sleep aids peak in the first three hours and have worn off by the time a sleep-maintenance patient wakes at 3am. The cannabis industry version of this mistake is the CBN-for-falling-asleep marketing claim, which the rigorous trial evidence does not support. CBN works for sleep maintenance, where its slower clearance is an advantage. For sleep onset, the molecule with the supporting evidence is THC at 5 to 10mg.
The job of this page is to give you the diagnostic frame, walk through what the chronic insomnia literature actually shows about cannabis, and then route you to the right protocol for the right subtype. This is the umbrella. The specific protocols live on the subtype pages.
What is chronic insomnia clinically, and when does it warrant treatment?
The threshold used in clinical research and by the American Academy of Sleep Medicine is three or more nights per week of sleep difficulty, lasting at least three months, with daytime consequences. The daytime consequence requirement matters. Many people have nights where sleep is poor without crossing into the territory that needs intervention. The clinical definition reserves "chronic insomnia" for sleep difficulty that meaningfully degrades waking function: persistent fatigue, mood disruption, concentration problems, irritability, or measurable performance issues at work or in daily tasks.
Roughly 10 to 15 percent of adults meet the chronic insomnia threshold in any given year, with prevalence rising with age and in women. The condition is comorbid with depression, anxiety, chronic pain, and PTSD in a majority of cases, which is part of why the treatment question is hard. Treating insomnia as a primary diagnosis works when the insomnia is the primary problem. When the insomnia is downstream of pain or anxiety, the primary condition needs treatment too, and a cannabinoid protocol targeted at sleep alone will produce thin and short-lived results.
The first-line treatment recommended by sleep medicine guidelines is not pharmacological. Cognitive behavioral therapy for insomnia (CBT-I) outperforms every drug class studied in head-to-head trials, with durable effects after the treatment ends. This is the awkward truth that most cannabis-and-sleep content avoids. Cannabis edibles are a reasonable second-line option, particularly for patients who have already tried CBT-I or who cannot access it, and for patients on prescription sleep medication who want to switch to something with a different side-effect profile. Cannabis is not the first thing to try.
What does the research show about cannabis for chronic insomnia?
The strongest current synthesis is the Suraev 2020 systematic review in Sleep Medicine Reviews, which pooled preclinical and clinical studies on cannabinoids and sleep. The conclusion was cautiously positive for THC at modest doses (sleep onset latency reduced by roughly 10 to 20 minutes versus placebo in controlled trials) and inconclusive for CBD as a sleep aid in isolation. The Babson 2017 Current Psychiatry Reports review reached similar conclusions and added the important caveat that long-term THC use can suppress REM sleep, which is the sleep stage most associated with memory consolidation and emotional processing. Short-term symptom relief is not the same as long-term sleep quality.
The most rigorous individual trial is Walsh 2021 in Sleep, a randomized crossover study of medicinal cannabis (a 1:1 THC:CBD nano-formulation) in 24 patients with chronic insomnia. Patients on the active product showed a 30-minute reduction in time-to-fall-asleep versus placebo, with no measurable change in total sleep time or sleep architecture. The trial was small but well-controlled and provides the cleanest evidence we have that THC-containing edibles can shorten sleep onset latency in this specific population.
For sleep maintenance alone, the Bonn-Miller 2024 placebo-controlled CBN trial in Experimental and Clinical Psychopharmacology is the most important paper in the field. The trial enrolled 293 adults with self-reported sleep disturbance and tested CBN at 25mg, 50mg, and 100mg against placebo over seven nights. The 25mg and 50mg doses produced statistically significant reductions in nighttime awakenings and improved subjective sleep quality. The 100mg dose did not outperform the 50mg dose, suggesting a ceiling effect. Critically, the trial measured both onset and maintenance, and found that CBN did not significantly improve sleep onset latency. The benefit was for staying asleep, not for falling asleep. This is the trial that justifies the editorial position EdibleRank takes across the Sleep cluster: CBN for maintenance, THC for onset.
Vigil 2018 in Medicines is the largest observational dataset on cannabis flower for insomnia (n=409), with patients tracking sleep effects across 1,056 sessions using a mobile app. Mean perceived insomnia symptom reduction was 4.5 points on a 0-to-10 scale. The observational design limits inference, but the magnitude of effect is consistent with the controlled trial data and the sample size is meaningful.
What none of these papers support is the idea that any single cannabinoid is a universal insomnia solution. Each one addresses a specific subtype of the condition at a specific dose. The dispensary aisle is full of products marketed as "sleep gummies" that combine CBN at sub-therapeutic doses with token THC, designed to be sold to cannabis-naive consumers who have not asked the diagnostic question this page is built around. Those products are not wrong. They are non-specific. They will work for some sleepers some of the time and fail for others, and the failure pattern will track exactly with which subtype the failed sleeper has.
How do you figure out which subtype you have?
Track three nights of sleep. Not seven. Three is enough.
Note the time you got into bed with the intention of sleeping. Note the time you estimate you fell asleep. Subtract. If the answer is more than 30 minutes on at least two of the three nights, you have a sleep-onset problem.
Note any awakenings during the night, the approximate time, and how long it took to fall back asleep. If you wake at least once per night between 2 and 5am and spend 30 minutes or more awake before returning to sleep, you have a sleep-maintenance problem.
Many chronic insomnia patients have both. In that case the dominant subtype (the one that bothers you more, or accounts for more lost sleep time) gets the primary protocol. The secondary subtype gets layered on later if the primary fix does not fully resolve the issue. The two protocols are compatible, but introducing them simultaneously makes it impossible to know which intervention is producing which result.
Do not use a sleep-tracking app or wearable as the only source of this data. Wearable sleep staging is not accurate enough to drive treatment decisions, and the false precision is worse than no data. A handwritten note on your phone is sufficient.
Protocol overview, with detailed protocols on the subtype pages
If your dominant subtype is sleep onset, the protocol is 5 to 10mg of THC in an edible taken 60 to 90 minutes before bed, ideally paired with a small amount of CBD (5mg or so) to soften the THC peak. CBN can be added but is not the primary driver of effect for this subtype. The full protocol with product picks and dosing schedule is at cannabis edibles for sleep onset.
If your dominant subtype is sleep maintenance, the protocol is 20mg of CBN taken 30 to 45 minutes before bed, with optional low-dose THC (2 to 5mg) as a sleep-onset support if you also have mild onset issues. The full protocol with product picks and dosing schedule is at cannabis edibles for sleep maintenance.
For both subtypes, the same supporting principles apply. Edibles peak at roughly 90 minutes after ingestion for traditional formulations and 45 to 60 minutes for nano-emulsion formulations. The bedtime dosing schedule needs to account for that lag. Tolerance builds at higher doses faster than at lower doses, which means staying at the minimum effective dose extends the useful life of the protocol. Two-week tolerance breaks every two to three months reset cannabinoid receptor sensitivity and recover the original effect at the original dose. And cannabis-related sleep effects diminish if the same product is taken every night without variation, which means rotating between two products with similar but not identical profiles often extends the useful life of the protocol beyond what either product alone would provide.
What is the long-term picture for cannabis as a sleep treatment?
The honest answer is that the long-term picture is not well-studied. Most controlled trials run for weeks, not years. The Babson 2017 review flagged THC-related REM suppression as a real concern for nightly long-term users, and subsequent observational data has reinforced that worry without producing a clear quantitative answer about how much REM suppression is too much. CBN does not appear to suppress REM at the doses used clinically, but the long-term data on CBN alone is thinner than for THC because CBN as a commercial cannabinoid is much newer.
The practical implication is that nightly use of any cannabinoid sleep aid for years should be discussed with a sleep medicine specialist, not adopted as an indefinite default. For many patients, cannabis edibles are the right tool for a defined period (six months to two years) during which CBT-I, addressing comorbid pain or anxiety, or other interventions can take effect. After that window, the goal should be tapering down rather than scaling up. The protocol that works is the protocol that helps you sleep tonight while you build the conditions that will help you sleep without it long-term.
Cannabis works for chronic insomnia. The trial evidence is real. The dependency risk is also real and underdiscussed in most cannabis content. The product recommendations on the subtype pages reflect that posture.