BakedIn.co

Stress & Sleep · Research-based

Sleep Restriction Therapy

Sleep Restriction Therapy (Spielman, Saskin & Thorpy, Sleep 1987;10(1):45-56) is the most counterintuitive of the CBT-I components and one of the most effective. The AASM 2021 CBT-I Clinical Practice Guideline rates it Grade A as a recommended component. The protocol shrinks your time in bed to closely match the actual sleep you have been getting, building homeostatic sleep pressure so that the time in bed becomes consolidated sleep rather than fragmented wakefulness. The practice itself is short — calculation and commitment. The hard part lives in your actual schedule for the following weeks. Always run a two-week sleep diary first so the calculation is real.

Evidence basis

Spielman AJ, Saskin P, Thorpy MJ. Treatment of chronic insomnia by restriction of time in bed. Sleep 1987;10(1):45-56; Edinger JD, Arnedt JT, Bertisch SM et al. Behavioral and psychological treatments for chronic insomnia disorder: AASM clinical practice guideline. J Clin Sleep Med 2021;17(2):255-262 — Grade A component; Miller CB et al., The evidence base of sleep restriction therapy for treating insomnia disorder. Sleep Med Rev 2014;18(5):415-424

Duration

5 min

Posture

Sitting

Difficulty

Intermediate

Format

Scripted

Benefits

InsomniaSleep onsetSleep efficiency

The practice

Step by step

  1. 01

    Sit at a table with two weeks of completed sleep diary entries in front of you. If you have not kept a diary for at least two weeks, stop and run the CBT-I Sleep Diary practice first — without baseline data this protocol will be guessing.

  2. 02

    Step 1 — Calculate average Total Sleep Time. Add up the TST values from the last 14 days of diary entries. Divide by 14. Write the result in hours and minutes — e.g., 'average TST = 5 hours 30 minutes.'

  3. 03

    Step 2 — Set your time-in-bed (TIB) window. The window equals your average TST plus 15 minutes — no more. So a 5-hour-30-minute TST gives a 5-hour-45-minute TIB. Important guardrail: never set TIB below 5 hours 30 minutes regardless of what TST shows. Going below that threshold is unsafe for daytime alertness and driving.

  4. 04

    Step 3 — Choose your fixed wake time. Pick the time you must be up most days — work alarm, family schedule, the time you would naturally rise without an alarm. This is your anchor; it does not move, including weekends. Write it down.

  5. 05

    Step 4 — Calculate your earliest bedtime. Subtract the TIB window from your fixed wake time. If wake = 6:00 a.m. and TIB = 5:45, earliest bedtime = 12:15 a.m. Write this down. You cannot go to bed before this time — even if you feel sleepy at 9 p.m.

  6. 06

    Step 5 — Commit aloud. Read your three numbers: 'I will go to bed no earlier than [time], get up at [wake time], regardless of how I sleep, for the next seven days.' Saying it aloud is not optional — it is the commitment step that separates a calculation from a behavior change.

  7. 07

    Step 6 — Plan the hardest part: staying up until the new late bedtime. Identify two low-stimulation activities for the evening — read a book in a dim room, listen to calm audio, do quiet handwork. Avoid screens, exercise, food after 8 p.m., and the bed itself until your scheduled bedtime.

  8. 08

    Step 7 — Run the week. Get into bed at the bedtime, get up at the wake time, no naps. Continue the sleep diary every morning. Expect the first 3-5 days to feel worse — daytime sleepiness is the homeostatic pressure building. Do not drive if dangerously sleepy.

  9. 09

    Step 8 — Weekly re-evaluation. After seven days, calculate your weekly average Sleep Efficiency (SE) from the diary. If SE ≥ 85%, expand the TIB window by 15 minutes (move bedtime 15 minutes earlier; keep wake time fixed). If SE is 80-85%, hold the current window for another week. If SE < 80%, contract the window by 15 minutes.

  10. 10

    Step 9 — Continue weekly adjustments. Most people reach a stable, consolidated sleep pattern within 4-8 weeks. The goal is not maximum sleep — it is sleep that is consolidated, efficient, and predictable. Stop tightening once you are sleeping continuously through your TIB window.

  11. 11

    Step 10 — Maintain the wake time even after the protocol ends. The fixed wake anchor is the single highest-yield habit from CBT-I, and the gains erode quickly if the wake time drifts more than an hour day to day. Bedtime can flex; wake time should not.

Modifications

Variations

  • Provider-supervised version (recommended for severe insomnia): a CBT-I clinician calibrates the TIB window using your diary, monitors daytime sleepiness with a validated clinical sleepiness measure, and adjusts weekly. If your insomnia has lasted more than three months and is affecting daytime function (per ICSD-3 chronic insomnia criteria), prefer this version. Free providers exist via the VA CBT-i Coach app and many telehealth platforms.

  • Time-in-bed compression instead of restriction (gentler): for people who cannot tolerate the standard protocol — older adults, people with significant medical comorbidity, those at high fall risk from sleep deprivation — set TIB to average TST plus 30-45 minutes rather than 15. Slower progress, lower daytime risk.

  • Pairs-with reminder: this practice almost always runs alongside Bootzin Stimulus Control (get out of bed if awake more than 20 minutes). They are complementary, not redundant — one sets the window, the other sets the rule for what happens inside it.

Note

Sleep Restriction Therapy is the highest-risk CBT-I component and the one most often misapplied. Absolute contraindications: bipolar disorder (acute sleep deprivation can trigger mania — requires psychiatric supervision); seizure disorder (sleep deprivation lowers seizure threshold); current pregnancy in the second or third trimester; untreated obstructive sleep apnea (restriction worsens apnea-driven fragmentation); occupations requiring sustained alertness with no margin for error (commercial driving, surgery, aviation) during the initial 2-3 weeks. Relative contraindications requiring clinician oversight: history of falls, significant cardiovascular disease, current use of sedative-hypnotics, age 75+ at high frailty. Do not run this protocol from this practice alone if any contraindication applies — bring the calculation to a CBT-I provider. Even for low-risk users, do not drive during periods of significant daytime sleepiness, and do not set TIB below 5 hours 30 minutes regardless of diary data.

Goes well with

Pairs with

← Back to practices