What Is the 3-2-1 Rule for Sleep? What the Evidence Says About Each Step

The 3-2-1 rule, and its longer version, 10-3-2-1-0, are popular bedtime rules. They are easy to remember, but they are not an official guideline. This page explains each step and checks it against the research. Some steps hold up well; others have little or no evidence behind them.

Quick Answer (TL;DR)

The 3-2-1 rule means: stop eating (and drinking alcohol) 3 hours before bed, stop working 2 hours before bed, and stop using screens 1 hour before bed. The 10-3-2-1-0 version adds no caffeine in the 10 hours before bed and no snooze button in the morning. It is a rule of thumb, not a clinical guideline, and no study has tested it as a whole. Of its steps, cutting caffeine well before bed has the strongest evidence.

Key Takeaways

• Caffeine: strong evidence. A 2023 meta-analysis of 24 studies found caffeine cut total sleep by 45 minutes. The authors advise having a standard coffee at least 8.8 hours before bed (Gardiner et al., 2023).
• Screens: some evidence. In a controlled study, reading on a light-emitting e-reader before bed, instead of a printed book, delayed sleep and the body clock (Chang et al., 2015).
• Late meals: mixed evidence. Observational studies mostly link eating close to bedtime with poorer sleep, but experiments are inconsistent (scoping review).
• Work: no direct studies of a 2-hour cut-off.
• No snooze: conflicting evidence. One small lab study found snoozing prolonged grogginess after waking (Ogawa et al., 2022); another found 30 minutes of snoozing did no clear harm in habitual snoozers (Sundelin et al., 2024).

What Is the 3-2-1 Rule? Each Step at a Glance

Is the 3-2-1 Rule Backed by Science?

Not as a package. We found no health body that publishes the rule, and no study that has tested it from start to finish. The individual steps can be checked against research, one at a time.

10: No caffeine 10 hours before bed (strong evidence)

This is the best-supported step. A 2023 meta-analysis of 24 studies found that caffeine reduced total sleep time by 45 minutes and sleep efficiency by 7%. It also added 9 minutes to the time it took to fall asleep and 12 minutes of time awake after falling asleep. It reduced deep sleep and increased light sleep. To avoid losing sleep, the authors concluded that a standard coffee (about 107 mg of caffeine) should be drunk at least 8.8 hours before bed, and a typical pre-workout supplement (217.5 mg) at least 13.2 hours before (Gardiner et al., 2023).

An earlier controlled study gave people 400 mg of caffeine at bedtime, or 3 or 6 hours before it. Even the dose taken 6 hours before bed disrupted sleep compared with a placebo (Drake et al., 2013).

The NHS advises no tea or coffee for at least 6 hours before bed (NHS). The right gap depends on the dose: for a standard coffee, the rule’s 10 hours is a little longer than the 8.8 hours the meta-analysis suggests. More: how afternoon coffee affects deep sleep.

3: No food 3 hours before bed (mixed evidence)

The evidence here is less clear-cut. A systematic scoping review found that most observational studies link eating close to bedtime with poorer sleep quality. Experimental studies, which can test cause and effect, were inconsistent, and differences between studies prevented a consensus (scoping review).

For people with acid reflux, there is more specific advice. The American College of Gastroenterology suggests avoiding meals within 2 to 3 hours of bedtime, a conditional recommendation based on low-quality evidence (ACG guideline, 2022). The NHS’s general insomnia advice is not to eat a big meal late at night (NHS).

3: No alcohol 3 hours before bed (some evidence)

Some versions of the rule include alcohol in the 3-hour step. A 2013 review of studies in healthy volunteers found that alcohol shortens the time it takes to fall asleep and makes the first half of the night more settled, but increases sleep disruption in the second half. It also delays the first period of REM sleep at all doses (Ebrahim et al., 2013). The review’s methods were later disputed by other researchers (Pressman et al., 2015). The NHS advises no alcohol for at least 6 hours before bed (NHS). More: why alcohol ruins your sleep.

2: No work 2 hours before bed (no direct studies)

No research has tested a 2-hour cut-off for work. The closest official advice is broader: the NHS recommends relaxing for at least an hour before bed, for example by taking a bath or reading a book (NHS).

1: No screens 1 hour before bed (some evidence)

In a randomized inpatient study, people read either on a light-emitting e-reader or a printed book before bed. The e-reader suppressed the evening rise in melatonin, delayed the body clock, lengthened the time it took to fall asleep and reduced alertness the next morning (Chang et al., 2015). Other researchers questioned how well the lab conditions reflect everyday screen use, and the authors replied (PubMed).

In children and teenagers, a 2016 meta-analysis linked having or using a portable device at bedtime with not getting enough sleep, poorer sleep quality and daytime sleepiness (Carter et al., 2016). These are observational findings, so they show a link, not proof of cause. The NHS advises not watching TV or using devices such as smartphones right before bed (NHS).

0: No snooze button (conflicting evidence)

The “0” step tells you not to hit snooze. Only a few lab studies have tested snoozing, and two small ones disagree:

• Against snoozing: in a sleep-lab study of 10 university students, using a snooze alarm prolonged sleep inertia, the groggy feeling after waking, compared with a single alarm (Ogawa et al., 2022).
• No clear harm: in 31 people who habitually snooze, 30 minutes of snoozing improved or did not affect performance on thinking tests right after getting up, compared with waking at the first alarm. It had no clear effect on morning sleepiness, mood, stress-hormone (cortisol) response or overnight sleep (Sundelin et al., 2024).

The studies differ in who took part and how long the snoozing lasted, and both are small. The honest summary is that the evidence on snoozing is not settled either way.

Should You Follow the 3-2-1 Rule?

Judged step by step:

• The caffeine and screen steps line up with official advice and research.
• The food and alcohol steps partly do. Official advice uses a longer gap for alcohol (6 hours) and avoids “big” late meals rather than all food.
• The work step is untested, though it fits the general advice to relax before bed.
• The no-snooze step has conflicting evidence from two small studies.

For other evidence-based routines, see how to spend the last hour before bed and daytime habits that help you fall asleep faster. If you’re lying awake after trying everything, see whether you should stay in bed if you can’t sleep.

Tracking Your Own Nights

If you try the rule, or just one step of it, a record of your nights lets you compare before and after rather than relying on how you feel.

With Snollo, you tap Start at bedtime and the app records the night on your iPhone. The free tier keeps your full sleep history, so you can compare nights over weeks. Sleep stages are most accurate with an Apple Watch. Without one, Snollo estimates approximate stages from the iPhone’s microphone and motion sensors, with the iPhone on the mattress near you. Phone and watch stages are estimates; the most accurate way to measure sleep stages is an overnight sleep study (polysomnography), which records brain waves (EEG), eye movements (EOG) and muscle tone (EMG) (NHLBI). Snollo is not a medical device and does not diagnose any condition.

Download Snollo free on the App Store

Frequently Asked Questions

What is the 3-2-1 rule for sleeping?

It is a bedtime rule of thumb: stop eating (and drinking alcohol) 3 hours before bed, stop working 2 hours before bed, and stop using screens 1 hour before bed. It is not an official guideline and has not been studied as a whole, but some of its steps match official advice.

What is the 10-3-2-1-0 rule?

An extended version of the 3-2-1 rule. It adds no caffeine in the 10 hours before bed and no hitting the snooze button in the morning (the 0). The caffeine step has the strongest research support of all the steps; for the no-snooze step, the two small lab studies on snoozing reached different conclusions.

Is the 3-2-1 sleep rule proven?

No. No study has tested the rule as a package. The evidence differs by step: strong for avoiding caffeine well before bed, some for screens and alcohol, mixed for late meals, and none specifically for a 2-hour work cut-off.

How long before bed should you stop drinking coffee?

A 2023 meta-analysis concluded that a standard coffee (about 107 mg of caffeine) should be drunk at least 8.8 hours before bedtime to avoid losing sleep. The NHS advises no tea or coffee for at least 6 hours before bed. Larger doses need a longer gap.

Is hitting snooze bad for you?

The evidence is limited and mixed. A lab study of 10 students found that using a snooze alarm prolonged grogginess after waking compared with a single alarm. A larger lab study of 31 people who habitually snooze found that 30 minutes of snoozing improved or did not affect thinking performance right after getting up, with no clear effect on morning sleepiness or mood. Both studies are small.

Snollo does not diagnose or treat any medical condition. This article summarizes published research and official advice for general information. If you have ongoing trouble sleeping, talk to a doctor. If you have any doubts about how you feel or how you sleep, talk to a doctor or sleep specialist.

Sources

• The effect of caffeine on subsequent sleep: systematic review and meta-analysis (Gardiner et al., 2023) — Leeds Beckett Repository
• Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed (Drake et al., 2013) — JCSM
• Chrono-nutrition and sleep: systematic scoping review — PMC, NIH
• ACG Clinical Guideline for GERD (2022) — PMC
• Alcohol and Sleep I: Effects on Normal Sleep (Ebrahim et al., 2013) — Alcohol Clin Exp Res
• Alcohol and Sleep Review: Flawed Design, Methods, and Statistics Cannot Support Conclusions (Pressman et al., 2015) — Alcohol Clin Exp Res
• Evening use of light-emitting eReaders negatively affects sleep (Chang et al., 2015) — PNAS
• Portable screen-based media device access or use and sleep outcomes: meta-analysis (Carter et al., 2016) — JAMA Pediatrics
• Effects of using a snooze alarm on sleep inertia after morning awakening (Ogawa et al., 2022) — PMC, NIH
• Is snoozing losing? (Sundelin et al., 2024) — PubMed
• Insomnia — NHS
• Sleep Studies — NHLBI