Specific Conditions · Anytime
CBT-I — The 4-Week Protocol Overview
CBT-I is the American College of Physicians' 2016 first-line treatment for chronic insomnia — recommended before any sleep medication. This entry orients you to the full 4-6 week protocol: what each component does, what to expect week by week, and how to find a trained provider. It's built for adults who have cycled through sleep hygiene tips for years and are wondering whether medication is their only remaining option.
Evidence basis
ACP 2016 Clinical Practice Guideline (first-line recommendation for chronic insomnia, ahead of pharmacotherapy); Bootzin stimulus control, 1972; Morin & Espie CBT-I protocol lineage; Espie et al. Sleepio RCT, Lancet Psychiatry 2019; Carney & Manber cognitive model; Czeisler chronobiology, Harvard (light/anchor timing); Drake et al., Wayne State (shift-work insomnia); Somryst FDA De Novo clearance 2020
Duration
30 min
When
Anytime
Level
Intermediate
Format
Protocol
Benefits
The protocol
Step by step
- 01
Before starting, complete a 7-day sleep diary: log your estimated time in bed, time to fall asleep, number of awakenings, and final wake time each morning. This baseline is not optional — sleep restriction dosing in Week 1 is calculated directly from it.
- 02
Calculate your average total sleep time (TST) from the diary. In Week 1, set your prescribed time-in-bed window to match that TST, with a hard floor of 5.5 hours minimum. If your diary shows you sleep 5 hours but spend 9 in bed, your initial window is 5.5 hours — not 9.
- 03
Set a fixed, non-negotiable rise time and work backward from it to establish your sleep window. The rise time anchors the entire protocol. Choose a time you can hold on workdays and weekends alike — even after a bad night.
- 04
Apply stimulus control starting Day 1 (Bootzin, 1972): use the bed only for sleep and sex. If you are not asleep within roughly 20 minutes, or if you wake in the night and cannot return to sleep, get out of bed and go to a dim, quiet space. Return only when you feel genuinely sleepy — not just tired, not just bored.
- 05
Expect Weeks 1-2 to be the hardest part of the protocol. Sleep restriction deliberately builds homeostatic sleep pressure by limiting time in bed to actual sleep time. You will feel more tired than usual during the day. This is the mechanism working, not a sign the protocol is failing.
- 06
Track your sleep efficiency each morning: divide total sleep time by total time in bed and multiply by 100. When your sleep efficiency reaches 85% or above for five consecutive nights, expand your sleep window by 15 minutes — add it to your bedtime, not your rise time.
- 07
In Weeks 2-3, begin cognitive restructuring. Write down the specific thoughts that keep you awake or that you rehearse about sleep — 'I'll be useless tomorrow,' 'I haven't slept properly in weeks,' 'Something is wrong with me.' Then apply structured examination: what is the actual evidence for and against this thought? What would you tell a colleague who said this to you?
- 08
Challenge catastrophic sleep-loss beliefs directly using the Morin cognitive model: sleep deprivation from one or two poor nights is uncomfortable but not dangerous for otherwise healthy adults. Performance impairment is real but typically overestimated by the sleep-anxious brain. Write this down; do not just think it.
- 09
Add a relaxation component in Week 2 or 3 — progressive muscle relaxation, diaphragmatic breathing, or body scan. These are not the primary mechanism of CBT-I; they reduce pre-sleep arousal so that sleep restriction and stimulus control can do their work. Pick one method and practice it daily, not only on bad nights.
- 10
Address sleep hygiene as a supporting layer, not the treatment itself: consistent light exposure in the morning (Czeisler chronobiology, Harvard), caffeine cutoff at least 6 hours before your sleep window, and a wind-down period of 30-60 minutes with reduced screen brightness. Sleep hygiene alone does not resolve chronic insomnia — it removes obstacles so the behavioral work can take effect.
- 11
By Week 4, most people with primary insomnia see sleep efficiency above 85%, reduced sleep-onset latency, and fewer middle-of-the-night awakenings. If you are not seeing improvement by the end of Week 4, this is a signal to seek a CBT-I-trained clinician — not a signal to add medication without behavioral work.
- 12
To find a trained provider, search the Society of Behavioral Sleep Medicine (SBSM) directory at behavioralsleep.org. Digital CBT-I programs (Sleepio, Somryst — FDA-cleared) have demonstrated efficacy in randomized trials and are a viable option when in-person access is limited.
- 13
Do not use this protocol as a reason to delay medical evaluation if you have symptoms of sleep apnea (witnessed apneas, loud snoring, waking with headaches, excessive daytime sleepiness despite adequate time in bed). CBT-I does not treat sleep-disordered breathing, and untreated apnea will undermine any behavioral protocol.
Modifications
Variations
Shift-worker adaptation — Your rise time anchor must match your actual schedule rotation. On a rotating shift, anchor to your earliest required wake time across the rotation and hold it even on days off. Sleep restriction is applied relative to your shifted sleep window, not a conventional night schedule. Drake et al. (Wayne State shift-work research) supports fixed-anchor approaches over free-running schedules during protocol.
Postpartum compressed version — Full sleep restriction is not appropriate when nighttime awakenings are infant-driven rather than insomnia-driven. Focus on stimulus control and cognitive restructuring only: use the out-of-bed rule for awakenings that are not infant-related, and work on catastrophic thoughts about fragmented sleep. Defer formal sleep restriction until the infant's sleep consolidates.
Perimenopausal adaptation — Night-sweat-driven awakenings complicate stimulus control because the trigger is physiological, not behavioral. Apply the out-of-bed rule only when you cannot return to sleep within 20 minutes after a sweat episode resolves. Cognitive restructuring targeting the 'I'll never sleep through the night again' belief is particularly high-yield in this population.
No-provider access workaround — If SBSM-listed providers are unavailable or unaffordable, Somryst (FDA 510k-cleared digital therapeutic) and Sleepio (peer-reviewed RCT evidence) deliver structured CBT-I digitally. These are not apps in the wellness sense; they are protocol-delivery tools with clinical trial backing.
Note
Sleep restriction is contraindicated in bipolar disorder due to documented mania-precipitation risk from sleep deprivation — do not self-administer this component without a clinician managing mood stability. Sleep restriction is also contraindicated in uncontrolled seizure disorders (sleep deprivation lowers seizure threshold) and active psychosis. Stimulus control is difficult to implement safely in severe anxiety disorders or PTSD where being out of bed alone in the dark is itself a trigger — a CBT-I-trained clinician can modify the protocol. This entry does not replace evaluation for comorbid sleep apnea, restless legs syndrome, or circadian rhythm disorders, all of which require separate assessment before or alongside CBT-I.