Caffeine and sleep

Key findings

Strong evidenceMechanistic · Lab study
Caffeine keeps you alert by blocking adenosine, a chemical that builds up while you are awake and makes you sleepy, so it hides sleep pressure rather than removing it.
Extracellular adenosine rises with time awake and falls during recovery sleep (animal microdialysis). Caffeine is an adenosine-receptor antagonist whose arousal effect depends on A2A receptors in mice. In humans it lowers slow-wave EEG activity, the marker of homeostatic sleep pressure.^1–5
Limits Receptor-level mechanism comes mainly from animal studies; human evidence is indirect (EEG and pharmacology).
Strong evidenceLab study · Guideline
It takes roughly 4 to 6 hours for the body to clear half of a caffeine dose, and this varies a lot from person to person.
Absorption is fast and complete: peak blood levels at about 30 min in fasted young men, 30–120 min across studies. EFSA gives a plasma half-life of about 4 h (range about 2–8 h). Individual studies in healthy adults report 2.7–9.9 h (n=10), 5.7 h (n=3) and 6.0 h in non-smokers (n=13).^6–10
Limits Small, older pharmacokinetic studies; most participants were young men.
Moderate evidenceLab study · Guideline
Smoking makes the body clear caffeine faster, while pregnancy, oestrogen-containing contraceptive pills and liver problems can make it linger much longer.
Mean half-life was 3.5 h in smokers vs 6.0 h in non-smokers, and 7.9 h on low-dose oestrogen pills vs 5.4 h in controls. It rose from about 3 h to about 10.5 h in the last 4 weeks of pregnancy (EFSA: 3–4 times longer at term). Two patients with alcoholic liver disease had half-lives of 60 h and 168 h.^6,8,9,11,12
Limits Small samples; liver data are two case observations. Dusk does not screen for these factors, so it should not tailor advice to them without asking.
Moderate evidenceLab study · Cohort
Genes affect how fast you break down caffeine and how sensitive your sleep is to it, but a single gene test does not set your personal cutoff.
CYP1A2 does about 95% of caffeine metabolism. The common -163C>A variant changed caffeine metabolism in smokers (n=51) but not in non-smokers (n=185). A variant of the adenosine A2A receptor gene (ADORA2A) is linked to self-rated caffeine sensitivity and to sleep-EEG responses, and this was replicated in a GWAS of 2,402 twins.^10,13–15
Limits Genotype explains only part of the variability; environment (smoking, hormones, drugs) matters as much.
Strong evidenceMeta-analysis · Systematic review
On average, caffeine taken later in the day shortens sleep by about three-quarters of an hour and makes it lighter and more broken.
A 2023 meta-analysis of 24 studies found total sleep time −45.3 min (95% CI 29.0–61.5), sleep efficiency −7%, sleep onset +9.1 min (3.8–14.4), wake after sleep onset +11.8 min (2.5–21.0) and deep N3 sleep −11.4 min (4.3–18.5), with light N1 sleep +6.1 min. REM duration and latency did not change significantly.^16,17
Limits Mostly healthy young adults with low-to-moderate habitual intake. Averages pool many doses and timings; effects at a given dose and time differ.
Moderate evidenceMeta-analysis · Randomised trial
The bigger the dose and the closer it is to bedtime, the more sleep caffeine tends to cost.
In meta-regression (k=30, n=262), sleep lost shrank by 2.8 min for each extra hour between caffeine and bed and grew by 0.2 min per extra mg. Five of six dose-comparison studies found 24–114 min less sleep with the larger dose. In a 12-person home study, 400 mg taken 6 h before bed still cut objectively measured sleep by more than 1 h.^16–18
Limits The linear model is a simplification; Drake 2013 had n=12 and used a headband monitor.
Moderate evidenceMeta-analysis
For a regular cup of coffee, research suggests stopping about 9 hours before bed to avoid losing sleep.
The meta-regression put the cutoff for coffee (107 mg per 250 mL) at 8.8 h before bedtime, meaning before 13:12 for a 22:00 bedtime. Later intake predicts a progressively larger loss of total sleep time.¹⁶
Limits The cutoff is based on statistical significance of total sleep time only; deep sleep may be affected earlier. Larger coffees (often 150–250+ mg) need more time.
Moderate evidenceMeta-analysis · Randomised trial
High-caffeine drinks such as pre-workouts and large energy drinks can affect sleep even when taken 12 or more hours before bed, so they are best kept to the morning.
The meta-regression cutoff for a standard pre-workout serve (217.5 mg) was 13.2 h before bed. In a randomised crossover trial (23 young men), 400 mg 12 h before bed delayed persistent sleep by 15 min and cut N3 by 21 min. At 8 h it added 29 min WASO and cut efficiency by 6.9%. At 4 h it cut total sleep by 51 min.^16,19
Limits The RCT studied men only, with habitual intake under 300 mg/day. The 400 mg dose is large (about four coffees).
Limited evidenceMeta-analysis · Randomised trial
Small amounts of caffeine, like a cup of tea, appear to have little measurable effect on sleep for most people.
The meta-regression found no cutoff time for a cup of black tea (about 47 mg). In a crossover RCT, 100 mg taken 12, 8 or even 4 h before bed had no significant effect on objective or subjective sleep.^6,16,19
Limits Young healthy men; sensitive people, older adults and people with insomnia may still be affected. EFSA notes 100 mg near bedtime may affect some adults.
Limited evidenceLab study
Even a morning coffee can leave a small trace in that night's sleep for some people.
In 9 healthy men, 200 mg at 07:10 still reduced total sleep time, sleep efficiency and slow-wave EEG power during sleep starting at 23:00, when saliva caffeine had fallen to 3 µmol/L.⁵
Limits A single small lab study; not a reason to tell most users to avoid morning coffee.
Moderate evidenceRandomised trial
People often don't notice how much caffeine is affecting their sleep, so feeling fine is not proof it has no effect.
With 400 mg 6 h before bed, the objective monitor showed more than 1 h less sleep, but the sleep diary showed −41 min (not significant, p=0.08). In a 2025 RCT the extra wake time after 400 mg at 4 and 8 h went unnoticed, and perceived quality dropped only in the 4-h condition.^18–20
Limits Small trials; the gap between perceived and measured sleep varies by person.
Moderate evidenceRandomised trial · Systematic review
Sleep tends to become more sensitive to caffeine from middle age onward.
With a high dose (400 mg), effects on sleep latency, duration and efficiency were larger in middle-aged adults (mean 52 y, n=24) than in young adults (mean 23 y, n=22). Caffeine also cut slow-wave sleep dose-dependently in both groups.^17,21,22
Limits Differences were clearest at high doses; at 200 mg effects were broadly similar across ages.
Mixed evidenceRandomised trial · Lab study
Regular coffee drinkers build some tolerance, but daily caffeine can still affect sleep.
At 400 mg three times daily for 7 days, sleep efficiency stayed reduced with only partial adaptation. Regular daytime intake (3 × 150 mg for 10 days) did not change total sleep or architecture in a late-night recording, but delayed REM sleep and left men feeling more tired on waking. Withdrawal after regular use raised evening sleepiness.^4,23–25
Limits Small, male-only samples; tolerance to sleep effects in real-world habitual users is not well quantified.
Limited evidenceLab study · Randomised trial
Caffeine in the evening may also push your body clock later, on top of keeping you awake.
About 200 mg (2.9 mg/kg) taken 3 h before habitual bedtime delayed the melatonin rhythm by about 40 min, roughly half the delay produced by 3 h of bright evening light. Regular morning-to-afternoon intake did not shift melatonin timing in another study.^25,26
Limits Only 5 participants in the evening study; the effect of afternoon caffeine on the clock is unclear.
Strong evidenceLab study · Guideline
The caffeine in a drink varies widely: tea usually has much less than coffee, and some energy drinks and pre-workouts have far more.
FDA figures per 12 oz: brewed coffee 113–247 mg, black tea about 71 mg, green tea about 37 mg, energy drinks 41–246 mg. Coffee-shop coffees ranged 58–259 mg per serve, and one drink varied 259–564 mg over six days. Brewed teas had 14–61 mg per serving. Pre-workouts averaged 254 mg per serve, and lab tests found 91–387 mg (59–176% of label).^27–31
Limits Brand, cup size and brewing method change content a lot; label claims for supplements can be inaccurate.
Strong evidenceGuideline
Up to about 400 mg of caffeine a day is considered safe for most healthy adults, but staying under that limit does not mean it is harmless for sleep.
EFSA and the FDA consider habitual intake up to 400 mg/day not a safety concern for healthy non-pregnant adults. EFSA also says single doses up to 200 mg are fine and pregnant women should keep to 200 mg/day. EFSA notes that a single 100 mg dose may lengthen sleep latency and shorten sleep in some adults, particularly close to bedtime.^6,27
Limits Safety limits do not apply to children and adolescents. People with heart conditions, anxiety or on interacting medicines should ask a clinician.
Moderate evidenceRandomised trial
For regular coffee drinkers, much of the morning lift comes from relieving the overnight dip that the caffeine habit itself creates.
In 369 adults, overnight abstinence in medium-to-high consumers made sleepiness and reaction time worse by the afternoon, and caffeine mostly restored them. In non/low consumers caffeine lowered sleepiness but raised jitteriness, with no net gain in mental alertness. A review concluded that net benefits under long-term use are small.^32–34
Limits Debated: caffeine clearly helps when people are sleep-deprived, and motor/physical benefits persist.
Moderate evidenceRandomised trial
Caffeine can help you push through a short night, but it does not make up for ongoing sleep loss.
During 5 nights limited to 5 h in bed (n=48), 200 mg caffeine twice daily kept vigilance up only for the first 3 days, and the caffeine group recovered more slowly afterwards. Reviews describe a cycle in which caffeine treats sleepiness that caffeine-disrupted sleep helped cause.^20,34,35
Limits Lab sleep restriction is more severe than typical daily life.
Limited evidenceCohort
Poor sleep and afternoon caffeine can feed each other: late caffeine worsens sleep, and a rough night can lead to more caffeine the next afternoon.
In 98 adolescents with 7 nights of at-home EEG, more caffeine (driven by afternoon and evening intake) predicted longer sleep onset and shorter, less efficient sleep. Lower sleep efficiency predicted more afternoon caffeine the next day. Most other sleep measures did not predict next-day intake.^20,36
Limits Observational, adolescent sample; the reverse direction (sleep to caffeine) was weak.
Limited evidenceLab study
A dip in alertness in the early afternoon is partly built into the body clock, and short or poor sleep makes it worse.
The post-lunch dip occurs even without lunch and without knowing the time. It is linked to circadian biology and is worsened by a high-carbohydrate lunch. In a driving simulator, sleepiness peaked mid-afternoon. 200 mg caffeine cut incidents to 34% of placebo, and caffeine plus a nap under 15 min cut them to 9%.^37,38
Limits Small lab studies; not everyone has a marked dip.
Strong evidenceSystematic review
Cutting caffeine suddenly can bring a few days of headache, tiredness or low mood, which usually pass within about a week.
Across 57 experimental and 9 survey studies, headache occurred in about 50%, and 13% had clinically significant distress or impairment. Symptoms began 12–24 h after the last dose, peaked at 20–51 h and lasted 2–9 days, appearing after stopping doses as low as 100 mg/day.³⁹
Limits Severity rises with usual daily dose; symptoms are not dangerous.
Limited evidenceRandomised trial · Guideline
Reducing caffeine gradually is a sensible way to avoid withdrawal, though moving your last cup earlier step by step has not been studied directly.
Manual-guided caffeine fading over 6 weeks reduced intake and distress, and the reduction held at 20 weeks. A remote RCT (n=109) cut intake from 565 to 233 mg/day and improved PROMIS sleep and anxiety scores. The FDA advises cutting back gradually. All of these taper the daily dose, not the timing.^27,39–41
Limits Trials enrolled people wanting to reduce problematic use. There is no evidence on tapering cutoff timing. Shifting the same total intake earlier should not itself cause withdrawal.
Mixed evidenceCohort
In everyday life the link between an evening coffee and that night's sleep is less clear than in the lab, partly because doses and sensitivity vary.
In 785 older adults (mean age 64; 5,164 nights of actigraphy plus diary), caffeine within 4 h of bedtime was not associated with that night's sleep duration, efficiency or wake after sleep onset, while evening alcohol and nicotine were.⁴²
Limits Doses were unknown, decaf may be included, sensitive people may self-select out of evening caffeine, and actigraphy cannot measure deep sleep.
Moderate evidenceGuideline
Sleep medicine guidance advises avoiding caffeine in the afternoon and evening, as one part of a broader plan rather than a cure on its own.
The AASM's patient guidance says to avoid caffeine in the afternoon or evening. The AASM 2021 insomnia guideline suggests not using sleep hygiene alone as treatment for chronic insomnia (conditional recommendation). Reviews find individual hygiene components relate to sleep but are largely untested as stand-alone public-health advice.^43–45
Limits People with chronic insomnia should be pointed to CBT-I rather than caffeine rules alone.

What experts and guidelines say

AASM (Sleep Education, "Healthy Sleep Habits", updated Aug 2020): "Avoid consuming caffeine in the afternoon or evening."⁴³. The AASM 2021 insomnia guideline also suggests clinicians not use sleep hygiene (which includes caffeine advice) as a single-component treatment for chronic insomnia (conditional)⁴⁴. Caffeine timing is a useful lever for poor sleepers, not a cure for insomnia⁴⁵.
EFSA NDA Panel (2015): 400 mg/day habitual and 200 mg single doses are safe for healthy non-pregnant adults; 200 mg/day in pregnancy; 100 mg near bedtime may affect sleep in some adults⁶.
FDA (Consumer Update, current as of Aug 2024): 400 mg/day for healthy adults; wide individual variation; cut back gradually to avoid withdrawal; pure and powdered caffeine is dangerous (toxic effects around 1,200 mg)²⁷.
Key researchers:
- Hans-Peter Landolt (Zurich): adenosine, caffeine, sleep EEG and genetics^4,5,14,17
- Timothy Roehrs, Thomas Roth and Christopher Drake (Henry Ford Hospital)^18,20
- Carissa Gardiner, Jonathon Weakley and Shona Halson (Australian Catholic University)^16,46
- Kenneth Wright Jr (Colorado): circadian effects²⁶
- Carolin Reichert, Janine Weibel and Christian Cajochen (Basel): chronic intake^4,24,25
- Julie Carrier (Montreal): ageing^21,22
- Laura Juliano and Roland Griffiths (withdrawal, caffeine use disorder)^39,40
- Peter Rogers (Bristol): withdrawal reversal^32,33
- Astrid Nehlig: caffeine pharmacology¹⁰
- Tarja Porkka-Heiskanen: adenosine as a sleep factor¹
- Timothy Monk: post-lunch dip³⁷
- Jim Horne: caffeine plus nap³⁸

Myths and overstatements

Partly true
“Caffeine doesn't affect me — I can have an espresso and still fall asleep.”
Sensitivity really does differ, partly for genetic reasons, and falling asleep is only delayed by about 9 min on average. Deep sleep, night-time wake and sleep efficiency can still be affected, and people often don't notice.^14,16,18,19
False
“Caffeine wears off after a few hours.”
With a half-life of about 4–6 h, roughly a fifth to a third of a coffee drunk 9 h before bed is still in the body at bedtime.^6,7
Overstated
“A fixed noon or 2 pm cutoff works for everyone.”
The safe gap depends on dose (tea little, regular coffee about 9 h, pre-workout about 13 h) and on the person (smoking, oral contraceptives, pregnancy, age).^{9,11,16,19,21}
Overstated
“Six hours between caffeine and bed is enough.”
This idea comes from Drake 2013, where 400 mg 6 h before bed still cut objective sleep by more than an hour. Pooled data point to about 9 h for a regular coffee and longer for high doses.^16,18
Overstated
“Coffee gives regular drinkers extra energy.”
In habitual users much of the lift reverses overnight withdrawal; net gains in mental alertness are small, though caffeine helps when you're sleep-deprived.^32–34
False
“You can run on caffeine instead of sleep.”
On 5-h nights, caffeine's benefit faded within about 3 days and recovery was slower afterwards.^20,35
Unproven
“Delay your first coffee 90–120 minutes after waking to prevent the afternoon crash.”
A popular podcast claim; we found no controlled trial testing it. The afternoon dip has a circadian basis, and in regular users a crash is more plausibly caffeine wearing off.^32,37
Partly true
“The afternoon slump is caused by eating lunch.”
The dip occurs even without lunch and is tied to the body clock; a high-carbohydrate lunch makes it worse.³⁷
Overstated
“Tea is as bad as coffee for sleep.”
A typical cup of tea has roughly half the caffeine of coffee or less, and the meta-regression found no cutoff time for black tea.^16,29
Overstated
“A CYP1A2 gene test tells you your personal caffeine cutoff.”
The common CYP1A2 variant mainly affects how far the enzyme can be induced (e.g. in smokers). Hormones, drugs, liver function and adenosine-receptor genes also matter.^10,13,14

References

1
Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW (1997). Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science.
Lab studydoi:10.1126/science.276.5316.1265 (opens in a new tab)
2
Huang ZL, Qu WM, Eguchi N, Chen JF, et al (2005). Adenosine A2A, but not A1, receptors mediate the arousal effect of caffeine. Nature Neuroscience.
Lab studydoi:10.1038/nn1491 (opens in a new tab)
3
Lazarus M, Shen HY, Cherasse Y, Qu WM, et al (2011). Arousal effect of caffeine depends on adenosine A2A receptors in the shell of the nucleus accumbens. Journal of Neuroscience.
Lab studydoi:10.1523/jneurosci.6730-10.2011 (opens in a new tab)
4
Reichert CF, Deboer T, Landolt HP (2022). Adenosine, caffeine, and sleep-wake regulation: state of the science and perspectives. Journal of Sleep Research.
Reviewdoi:10.1111/jsr.13597 (opens in a new tab)
5
Landolt HP, Werth E, Borbély AA, Dijk DJ (1995). Caffeine intake (200 mg) in the morning affects human sleep and EEG power spectra at night. Brain Research.
Lab study · n = 9doi:10.1016/0006-8993(95)00040-w (opens in a new tab)
6
EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA) (2015). Scientific Opinion on the safety of caffeine. EFSA Journal 13(5):4102.
Guidelinedoi:10.2903/j.efsa.2015.4102 (opens in a new tab)
7
Blanchard J, Sawers SJ (1983). The absolute bioavailability of caffeine in man. European Journal of Clinical Pharmacology.
Lab study · n = 10doi:10.1007/bf00613933 (opens in a new tab)
8
Statland BE, Demas TJ (1980). Serum caffeine half-lives. Healthy subjects vs. patients having alcoholic hepatic disease. American Journal of Clinical Pathology.
Lab study · n = 5doi:10.1093/ajcp/73.3.390 (opens in a new tab)
9
Parsons WD, Neims AH (1978). Effect of smoking on caffeine clearance. Clinical Pharmacology & Therapeutics.
Lab study · n = 26doi:10.1002/cpt197824140 (opens in a new tab)
10
Nehlig A (2018). Interindividual Differences in Caffeine Metabolism and Factors Driving Caffeine Consumption. Pharmacological Reviews.
Reviewdoi:10.1124/pr.117.014407 (opens in a new tab)
11
Abernethy DR, Todd EL (1985). Impairment of caffeine clearance by chronic use of low-dose oestrogen-containing oral contraceptives. European Journal of Clinical Pharmacology.
Lab study · 9 OC users + 9 controlsdoi:10.1007/bf00544361 (opens in a new tab)
12
Knutti R, Rothweiler H, Schlatter C (1982). The effect of pregnancy on the pharmacokinetics of caffeine. Archives of Toxicology Supplement.
Lab studydoi:10.1007/978-3-642-68511-8_33 (opens in a new tab)
13
Sachse C, Brockmöller J, Bauer S, Roots I (1999). Functional significance of a C-->A polymorphism in intron 1 of the cytochrome P450 CYP1A2 gene tested with caffeine. British Journal of Clinical Pharmacology.
Lab study · n = 236doi:10.1046/j.1365-2125.1999.00898.x (opens in a new tab)
14
Rétey JV, Adam M, Khatami R, et al (2007). A genetic variation in the adenosine A2A receptor gene (ADORA2A) contributes to individual sensitivity to caffeine effects on sleep. Clinical Pharmacology & Therapeutics.
Lab studydoi:10.1038/sj.clpt.6100102 (opens in a new tab)
15
Byrne EM, Johnson J, McRae AF, Nyholt DR, et al (2012). A genome-wide association study of caffeine-related sleep disturbance: confirmation of a role for a common variant in the adenosine receptor. Sleep.
Cohort · n = 2402doi:10.5665/sleep.1962 (opens in a new tab)
16
Gardiner C, Weakley J, Burke LM, et al (2023). The effect of caffeine on subsequent sleep: A systematic review and meta-analysis. Sleep Medicine Reviews.
Meta-analysis · 24 studiesdoi:10.1016/j.smrv.2023.101764 (opens in a new tab)
17
Clark I, Landolt HP (2017). Coffee, caffeine, and sleep: A systematic review of epidemiological studies and randomized controlled trials. Sleep Medicine Reviews.
Systematic reviewdoi:10.1016/j.smrv.2016.01.006 (opens in a new tab)
18
Drake C, Roehrs T, Shambroom J, Roth T (2013). Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. Journal of Clinical Sleep Medicine.
Randomised trial · n = 12doi:10.5664/jcsm.3170 (opens in a new tab)
19
Gardiner CL, Weakley J, Burke LM, et al (2025). Dose and timing effects of caffeine on subsequent sleep: a randomized clinical crossover trial. Sleep.
Randomised trial · n = 23doi:10.1093/sleep/zsae230 (opens in a new tab)
20
Roehrs T, Roth T (2008). Caffeine: sleep and daytime sleepiness. Sleep Medicine Reviews.
Reviewdoi:10.1016/j.smrv.2007.07.004 (opens in a new tab)
21
Robillard R, Bouchard M, Cartier A, Nicolau L, Carrier J (2015). Sleep is more sensitive to high doses of caffeine in the middle years of life. Journal of Psychopharmacology.
Randomised trial · n = 46doi:10.1177/0269881115575535 (opens in a new tab)
22
Drapeau C, Hamel-Hébert I, Robillard R, Selmaoui B, Filipini D, Carrier J (2006). Challenging sleep in aging: the effects of 200 mg of caffeine during the evening in young and middle-aged moderate caffeine consumers. Journal of Sleep Research.
Randomised trial · n = 24doi:10.1111/j.1365-2869.2006.00518.x (opens in a new tab)
23
Bonnet MH, Arand DL (1992). Caffeine use as a model of acute and chronic insomnia. Sleep.
Lab study · n = 12doi:10.1093/sleep/15.6.526 (opens in a new tab)
24
Weibel J, Lin YS, Landolt HP, et al (2021). Regular Caffeine Intake Delays REM Sleep Promotion and Attenuates Sleep Quality in Healthy Men. Journal of Biological Rhythms.
Randomised trial · n = 20doi:10.1177/07487304211013995 (opens in a new tab)
25
Weibel J, Lin YS, Landolt HP, et al (2020). Caffeine-dependent changes of sleep-wake regulation: Evidence for adaptation after repeated intake. Progress in Neuro-Psychopharmacology & Biological Psychiatry.
Randomised trial · n = 20doi:10.1016/j.pnpbp.2019.109851 (opens in a new tab)
26
Burke TM, Markwald RR, McHill AW, et al (2015). Effects of caffeine on the human circadian clock in vivo and in vitro. Science Translational Medicine.
Lab study · n = 5doi:10.1126/scitranslmed.aac5125 (opens in a new tab)
27
U.S. Food and Drug Administration (2024). Spilling the Beans: How Much Caffeine is Too Much?. FDA Consumer Update (content current as of 28 Aug 2024).
Guidelinefda.gov (opens in a new tab)
28
McCusker RR, Goldberger BA, Cone EJ (2003). Caffeine content of specialty coffees. Journal of Analytical Toxicology.
Lab studydoi:10.1093/jat/27.7.520 (opens in a new tab)
29
Chin JM, Merves ML, Goldberger BA, Sampson-Cone A, Cone EJ (2008). Caffeine content of brewed teas. Journal of Analytical Toxicology.
Lab study · 20 tea productsdoi:10.1093/jat/32.8.702 (opens in a new tab)
30
Jagim AR, Harty PS, Camic CL (2019). Common Ingredient Profiles of Multi-Ingredient Pre-Workout Supplements. Nutrients.
Cross-sectional · 100 productsdoi:10.3390/nu11020254 (opens in a new tab)
31
Desbrow B, Hall S, O'Connor H, Slater G, Barnes K, Grant G (2019). Caffeine content of pre-workout supplements commonly used by Australian consumers. Drug Testing and Analysis.
Lab study · 15 productsdoi:10.1002/dta.2501 (opens in a new tab)
32
Rogers PJ, Heatherley SV, Mullings EL, Smith JE (2013). Faster but not smarter: effects of caffeine and caffeine withdrawal on alertness and performance. Psychopharmacology.
Randomised trial · n = 369doi:10.1007/s00213-012-2889-4 (opens in a new tab)
33
James JE, Rogers PJ (2005). Effects of caffeine on performance and mood: withdrawal reversal is the most plausible explanation. Psychopharmacology.
Reviewdoi:10.1007/s00213-005-0084-6 (opens in a new tab)
34
McLellan TM, Caldwell JA, Lieberman HR (2016). A review of caffeine's effects on cognitive, physical and occupational performance. Neuroscience & Biobehavioral Reviews.
Reviewdoi:10.1016/j.neubiorev.2016.09.001 (opens in a new tab)
35
Doty TJ, So CJ, Bergman EM, et al (2017). Limited Efficacy of Caffeine and Recovery Costs During and Following 5 Days of Chronic Sleep Restriction. Sleep.
Randomised trial · n = 48doi:10.1093/sleep/zsx171 (opens in a new tab)
36
Lunsford-Avery JR, Kollins SH, Kansagra S, Wang KW, Engelhard MM (2022). Impact of daily caffeine intake and timing on electroencephalogram-measured sleep in adolescents. Journal of Clinical Sleep Medicine.
Cohort · n = 98doi:10.5664/jcsm.9736 (opens in a new tab)
37
Monk TH (2005). The post-lunch dip in performance. Clinics in Sports Medicine.
Reviewdoi:10.1016/j.csm.2004.12.002 (opens in a new tab)
38
Reyner LA, Horne JA (1997). Suppression of sleepiness in drivers: combination of caffeine with a short nap. Psychophysiology.
Randomised trial · n = 12doi:10.1111/j.1469-8986.1997.tb02148.x (opens in a new tab)
39
Juliano LM, Griffiths RR (2004). A critical review of caffeine withdrawal: empirical validation of symptoms and signs, incidence, severity, and associated features. Psychopharmacology.
Review · 57 experimental + 9 survey studiesdoi:10.1007/s00213-004-2000-x (opens in a new tab)
40
Sweeney MM, Meredith SE, Juliano LM, Evatt DP, Griffiths RR (2019). A randomized controlled trial of a manual-only treatment for reduction and cessation of problematic caffeine use. Drug and Alcohol Dependence.
Randomised trialdoi:10.1016/j.drugalcdep.2018.10.034 (opens in a new tab)
41
Knerr AW, Strickland JC, Labos BH, Lee DC (2026). Randomized controlled trial of a remote-guided caffeine reduction intervention. Drug and Alcohol Dependence.
Randomised trial · n = 109doi:10.1016/j.drugalcdep.2026.113123 (opens in a new tab)
42
Spadola CE, Guo N, Johnson DA, et al (2019). Evening intake of alcohol, caffeine, and nicotine: night-to-night associations with sleep duration and continuity among African Americans in the Jackson Heart Sleep Study. Sleep.
Cohort · n = 785doi:10.1093/sleep/zsz136 (opens in a new tab)
43
American Academy of Sleep Medicine (2020). 68% of Americans lose sleep to drinking alcohol, survey says. AASM press release (via Newswise).
Expert consensus · 2007 survey respondentsnewswise.com (opens in a new tab)
44
Edinger JD, Arnedt JT, Bertisch SM, et al (2021). Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. Journal of Clinical Sleep Medicine.
Guidelinedoi:10.5664/jcsm.8986 (opens in a new tab)
45
Irish LA, Kline CE, Gunn HE, Buysse DJ, Hall MH (2015). The role of sleep hygiene in promoting public health: a review of empirical evidence. Sleep Medicine Reviews.
Reviewdoi:10.1016/j.smrv.2014.10.001 (opens in a new tab)
46
Gardiner C, Weakley J, Burke LM, et al (2025). The effect of alcohol on subsequent sleep in healthy adults: a systematic review and meta-analysis. Sleep Medicine Reviews.
Meta-analysis · 27 studiesdoi:10.1016/j.smrv.2024.102030 (opens in a new tab)

Key findings

Caffeine keeps you alert by blocking adenosine, a chemical that builds up while you are awake and makes you sleepy, so it hides sleep pressure rather than removing it.

It takes roughly 4 to 6 hours for the body to clear half of a caffeine dose, and this varies a lot from person to person.

Smoking makes the body clear caffeine faster, while pregnancy, oestrogen-containing contraceptive pills and liver problems can make it linger much longer.

Genes affect how fast you break down caffeine and how sensitive your sleep is to it, but a single gene test does not set your personal cutoff.

On average, caffeine taken later in the day shortens sleep by about three-quarters of an hour and makes it lighter and more broken.

The bigger the dose and the closer it is to bedtime, the more sleep caffeine tends to cost.

For a regular cup of coffee, research suggests stopping about 9 hours before bed to avoid losing sleep.

High-caffeine drinks such as pre-workouts and large energy drinks can affect sleep even when taken 12 or more hours before bed, so they are best kept to the morning.

Small amounts of caffeine, like a cup of tea, appear to have little measurable effect on sleep for most people.

Even a morning coffee can leave a small trace in that night's sleep for some people.

People often don't notice how much caffeine is affecting their sleep, so feeling fine is not proof it has no effect.

Sleep tends to become more sensitive to caffeine from middle age onward.

Regular coffee drinkers build some tolerance, but daily caffeine can still affect sleep.

Caffeine in the evening may also push your body clock later, on top of keeping you awake.

The caffeine in a drink varies widely: tea usually has much less than coffee, and some energy drinks and pre-workouts have far more.

Up to about 400 mg of caffeine a day is considered safe for most healthy adults, but staying under that limit does not mean it is harmless for sleep.

For regular coffee drinkers, much of the morning lift comes from relieving the overnight dip that the caffeine habit itself creates.

Caffeine can help you push through a short night, but it does not make up for ongoing sleep loss.

Poor sleep and afternoon caffeine can feed each other: late caffeine worsens sleep, and a rough night can lead to more caffeine the next afternoon.

A dip in alertness in the early afternoon is partly built into the body clock, and short or poor sleep makes it worse.

Cutting caffeine suddenly can bring a few days of headache, tiredness or low mood, which usually pass within about a week.

Reducing caffeine gradually is a sensible way to avoid withdrawal, though moving your last cup earlier step by step has not been studied directly.

In everyday life the link between an evening coffee and that night's sleep is less clear than in the lab, partly because doses and sensitivity vary.

Sleep medicine guidance advises avoiding caffeine in the afternoon and evening, as one part of a broader plan rather than a cure on its own.

What experts and guidelines say

Myths and overstatements

“Caffeine doesn't affect me — I can have an espresso and still fall asleep.”

“Caffeine wears off after a few hours.”

“A fixed noon or 2 pm cutoff works for everyone.”

“Six hours between caffeine and bed is enough.”

“Coffee gives regular drinkers extra energy.”

“You can run on caffeine instead of sleep.”

“Delay your first coffee 90–120 minutes after waking to prevent the afternoon crash.”

“The afternoon slump is caused by eating lunch.”

“Tea is as bad as coffee for sleep.”

“A CYP1A2 gene test tells you your personal caffeine cutoff.”

References