The treatment that works best for long-term insomnia is not a tablet, and it is almost unknown outside sleep clinics. It starts by spending less time in bed.

Your sleep window

Three numbers and the arithmetic that a sleep clinic does on the first visit. Tick anything on the list that applies to you first: some of it means this protocol is not the right starting point, and the tool will say so instead of giving you a window.

What insomnia actually is

• It is not defined by hours. Insomnia disorder means difficulty falling asleep, staying asleep or waking too early, at least three nights a week for three months or more, plus a daytime cost — tiredness, mood, concentration, performance. Somebody who sleeps six hours and functions well does not have it. Somebody who sleeps seven and is wrecked may have something else entirely.
• The engine is hyperarousal. A body and brain that stay switched on: racing thoughts at lights-out, a jolt of alertness the moment the head hits the pillow, the clock arithmetic. Sleep is not something the nervous system does under pressure.
• Then conditioning takes over. After enough nights, the bedroom itself becomes a cue for wakefulness. That is learned, which is exactly why it can be unlearned, and it is the part sleeping tablets do nothing about.
• Trying harder makes it worse. Effort is arousal. This is the single most useful thing to understand about insomnia, and it is why the treatment removes the opportunity to lie there trying rather than adding relaxation on top.
• A bad night is not dangerous. Catastrophic thinking about the consequences of one bad night is itself one of the strongest maintainers of insomnia. Sleep is remarkably resilient; the anxiety about it is what persists.
• Waking in the small hours is normal; staying awake is the problem. Everyone surfaces several times a night, and the second half of the night is lighter sleep by design, so a 3 a.m. waking is not a fault. What turns it into insomnia is what happens next: checking the clock, calculating, and lying there trying. Alcohol in the evening makes exactly this waking far more likely.
• Short sleepers exist. A small number of people genuinely function on five or six hours. If daytime function is good, there is nothing to fix, whatever a wearable says about your score.

The treatment, in five parts

This is what cognitive behavioural therapy for insomnia consists of. Over three to eight weeks it outperforms sleeping tablets and, unlike them, the effect is still there a year later. It is usually delivered by a nurse, psychologist or a digital programme — and the two components doing most of the work are the two nobody would guess.

1. Sleep restriction — the engine. Match your time in bed to the sleep you are actually getting, so that sleep becomes concentrated and solid, then extend it fifteen minutes at a time. This is what the calculator above works out, and it is the part that feels wrong before it works.
2. Stimulus control — the retraining. Bed is for sleep and sex only. If you are awake and frustrated for what feels like around twenty minutes, get up, go to another room, do something dull in dim light, and return only when sleepy. Repeat as often as it takes. Do not clock-watch: turn the clock away.
3. A fixed rise time, seven days a week. Wake time is the anchor of the whole body clock, and lying in at weekends to catch up is the commonest way people undo their own progress. Bedtime can drift; getting-up time may not.
4. Cognitive work on the beliefs. “I need eight hours or I cannot function”, “tomorrow is ruined”, “I have never slept well” — these are testable predictions, and testing them is a therapeutic act rather than positive thinking.
5. A wind-down, which is the least important part and gets ninety per cent of the public attention. Dim light, no work, no arguments, no doom-scrolling in the hour before bed — useful, but it will not fix conditioned insomnia on its own.

Sleeping tablets, honestly

• They work, briefly. Z-drugs and benzodiazepines reduce the time to fall asleep by roughly a quarter of an hour and add a little total sleep. That is real, and it is a fraction of what people believe they are getting.
• They are licensed for days to a few weeks, and tolerance builds in about that time. Beyond it, most long-term users are taking them to avoid withdrawal insomnia rather than for a benefit — which feels identical from the inside.
• The costs are unglamorous and real: next-day slowness, roughly double the risk of falls and hip fractures in older people, impaired driving, memory effects, and dependence. Combining them with alcohol or opioids is the dangerous combination.
• Never stop them abruptly after long use. Withdrawal can mean rebound insomnia, agitation, and in some cases seizures. A slow, planned taper with a prescriber — ideally alongside CBT-I, which is what makes tapers succeed — is the route.
• Melatonin is not a sleeping tablet. It is a body-clock signal: modestly useful for jet lag, shift work and delayed sleep phase, and for some older adults, taken at the right time — usually a small dose a few hours before the target bedtime, not a large one at lights-out.
• Sedating antihistamines and alcohol are the two most common self-prescriptions, and both make sleep worse. Antihistamines produce tolerance within days and a heavy head; alcohol reliably destroys the second half of the night.

The thing to rule out before anything else

Obstructive sleep apnoea is common, badly under-diagnosed, and frequently mislabelled as insomnia or as ordinary tiredness. No amount of sleep restriction will fix breathing that stops sixty times an hour, and the untreated condition raises blood pressure, drives atrial fibrillation and multiplies road-crash risk.

The other things that masquerade as insomnia

• Restless legs: an urge to move the legs in the evening, relieved by moving, worse at rest. Treatable, often linked to low iron stores, and completely unaffected by sleep hygiene advice.
• Pain, reflux, itching and needing to pass urine — the boring physical causes of broken nights, each with its own treatment, and each routinely skipped over on the way to a sleeping tablet.
• Depression and anxiety run in both directions. Insomnia is often the first symptom and the last to leave, and treating the insomnia specifically improves the mood disorder as well — which is why it deserves its own treatment rather than being left to resolve on its own.
• Menopause, where hot flushes and night waking are a specific and treatable pattern rather than an inevitable phase.
• Thyroid disease, some medicines and caffeine metabolism. Steroids, some antidepressants, decongestants, beta-agonists and thyroid excess all fragment sleep — worth naming at a medication review.
• Circadian mismatch rather than insomnia: a body clock that runs late, forced to start work at seven. That is a timing problem, and light exposure and gradual shifts fix it where sleeping tablets cannot.

Sleep hygiene, sorted by how much it matters

Shift work, which is a different problem

• You are not failing at sleep hygiene; you are working against a body clock. Shift-work sleep disorder is its own diagnosis, and the goal is damage limitation rather than normal sleep.
• Anchor sleep helps: keeping one block of sleep at the same clock time whatever the rota, even if the rest moves, gives the body clock something stable to hold on to.
• Use light deliberately. Bright light early in the night shift, dark glasses on the commute home, and a dark, cool bedroom with the phone out of reach. Light is the strongest signal there is, and most shift workers get it exactly backwards.
• Caffeine at the start of the shift, not the end, and a planned nap before a night shift where the job allows one.
• The drive home is the dangerous part. Post-night-shift driving is one of the highest-risk things people do routinely: if you are fighting sleep, nap in the car park before you set off.
• Rotating forwards is easier than backwards — days to evenings to nights — and it is worth asking whether the rota can be built that way, because that is an organisational fix rather than a personal one.

Children and teenagers

• Teenagers are not lazy: their body clocks genuinely shift later at puberty, by an hour or two. A teenager who cannot sleep at ten and cannot wake at seven is displaying normal biology colliding with a timetable.
• Later school starts improve sleep, mood, grades and crash rates where they have been trialled, which makes this one of the few sleep problems that is genuinely a policy question rather than a parenting one.
• Consistency beats duration for children: the same routine and the same get-up time do more than an earlier bedtime imposed on an unwilling child.
• Phones in the bedroom are worth a fight, not because of blue light but because of the group chat, the game and the endless scroll — and because the notification at midnight wakes a sleeping child.
• Snoring in a child is not cute. Loud habitual snoring, pauses, mouth-breathing and daytime irritability can mean enlarged tonsils and sleep apnoea, which is treatable and is frequently misread as behavioural.
• For babies, this page is the wrong one — infant sleep runs on completely different rules, and safe-sleep guidance matters more than any technique.

If someone you love cannot sleep

• Do not say “just relax” or “have you tried not looking at your phone”. Both translate as “you are doing this to yourself” to somebody who has already tried everything the internet offers.
• Ask the apnoea questions, because the person who snores cannot hear themselves and a partner is often the only witness there will ever be. That question has diverted a great many people to the right diagnosis.
• Offer the name of the treatment. Most people have never heard of CBT-I and assume the choice is tablets or endurance. Knowing there is a third option is often the whole intervention.
• Support the fixed rise time, which is the hardest part to keep alone at weekends. Being the person who does not suggest a lie-in is genuinely helpful.
• Take the night driving seriously. Offer to drive, or to collect them, rather than debating whether they are too tired.
• If they are hopeless as well as sleepless, ask directly about mood and about thoughts of self-harm. Insomnia and depression travel together, and severe insomnia is an independent risk factor worth naming out loud.

The drill: 16 decisions

Sixteen ordinary nights and mornings — a bad week, a weekend lie-in, a snoring partner, a tablet that has lasted four years. Most have an instinctive answer that keeps the insomnia running. Pick your move; every answer explains why.

The card

Print it and put it where you will see it at two in the morning.

🖨️ Print the card