Bean Reaper
When to Drink Coffee on a Night Shift: A Caffeine Timing Guide
Share
Two people drink identical cups at 2 a.m. One is asleep by 8. The other watches the ceiling until noon.
Nothing about the coffee explains that. Everything about the timing and the person does — and both are things you can work with once you understand what caffeine is actually doing in there.
This is the mechanics piece. Half-life, the adenosine story, an hour-by-hour plan for a night shift, the coffee nap, when to stop, and where the evidence gets thin. Sources are linked throughout so you can check them.
What caffeine is actually doing
Caffeine is an adenosine receptor antagonist. Adenosine builds up in your brain across every waking hour and signals sleep pressure — the heaviness that grows the longer you have been awake. Caffeine fits the same receptor and occupies it, so the signal cannot land.
Read that again, because the crucial word is occupies. Caffeine blocks the receptor. It does not clear the adenosine.
Sleep pressure keeps accumulating behind the block for as long as you are awake. When caffeine clears, the whole backlog arrives at once. That is why the crash feels wildly out of proportion to the cup that came before it — you are not losing a stimulant, you are getting back a bill that has been quietly compounding.
This is also why caffeine cannot substitute for sleep. An Institute of Medicine review for the NIH found caffeine restored alertness in sleep-deprived subjects to only about 50% of the well-rested condition, with effects declining after roughly four and a half hours. The same review found 150 mg was about as effective as 300 or 600 mg. Past a fairly modest point, more caffeine simply does not buy more alertness.
The numbers you need
Onset
Peak plasma concentration arrives somewhere between 15 and 120 minutes after drinking. NIOSH's practical framing is that caffeine takes about 30 minutes to take effect, with a range of 15 to 45 minutes.
Thirty minutes is the number to plan around. It means the cup you drink when you already feel terrible is thirty minutes too late.
Half-life
In healthy adults, caffeine's half-life averages roughly five hours. NIOSH puts it at five to six and notes it can remain in your system considerably longer. Individual half-lives range roughly from under two hours to around ten.
That range is the single most important fact in this article. It means generic advice about cutoff times is close to useless, and it explains the two people at 2 a.m.
What moves it:
- Genetics. Variation in the CYP1A2 enzyme drives most person-to-person differences.
- Oral contraceptives. Roughly double the half-life.
- Pregnancy. Up to about 15 hours in the final trimester.
- Smoking. Shortens it by up to 50%.
Half-life math, illustrated
Using the average five-hour half-life and a single 200 mg dose, purely as an illustration:
| Hours since the cup | Roughly still in you |
|---|---|
| 0 | 200 mg |
| 5 | 100 mg |
| 10 | 50 mg |
| 15 | 25 mg |
| 20 | ~12 mg |
This is arithmetic, not a measurement of you. If your personal half-life is eight hours, every row shifts substantially. If it is three, the tail is far shorter. But it shows the shape of the problem: a cup at 3 a.m. is meaningfully present at 8 a.m., which is when a night-shift worker is trying to sleep.
How much is in a cup
Mayo Clinic gives 96 mg per 8 oz of brewed coffee and 63 mg per 1 oz espresso. The FDA gives 113 to 247 mg for a 12 oz brewed coffee.
Real-world variation dwarfs both. A survey of 97 commercial espressos found 25 to 214 mg per serve — and a separate Glasgow study found a range of 51 to 322 mg. When you buy coffee, you are not buying a known dose.
Roast level, meanwhile, barely matters. A 2024 study in Scientific Reports brewed filter coffee by weight across five roast levels and found the entire light-to-dark spread landed around 134–165 mg per 8 oz cup — a narrower range than the cup-to-cup variation within a single drink type. Dose, grind and brew method swamp roast level completely. Full breakdown in how much caffeine is in a cup of coffee.
The night shift timing table
This maps an 11 p.m. to 7 a.m. shift with a target sleep window starting around 9 a.m. Shift the whole thing to match your own hours and your own sensitivity.
| Clock time | Hours to sleep window | Strategy | Why |
|---|---|---|---|
| 10:30 p.m. (pre-shift) | ~10.5 | First cup, about 30 minutes before you need it | Onset is ~30 minutes; this gets you sharp as you clock in |
| 11 p.m.–1 a.m. | 8–10 | Hold. No extra caffeine. | You are not fighting the circadian low yet |
| 1–2 a.m. | 7–8 | Main cup of the shift. Coffee nap here if you can. | Enough runway for the tail to decay before bed |
| 3–4 a.m. | 5–6 | Last caffeinated cup, and keep it small — or skip it | At the average half-life, roughly half is still present at bedtime |
| 4–6 a.m. | 3–5 | Light, movement, cold water, decaf | NIOSH: caffeine near the end of a night shift causes restlessness the morning after |
| 6–7 a.m. | 2–3 | Nothing caffeinated | This cup is borrowed directly from your sleep |
| Drive home | ~2 | If drowsy: pull over, nap 15–20 minutes | NIOSH's explicit guidance for drowsy driving |
| Sleep window | 0 | Dark room, cool, phone elsewhere | Daytime sleep needs more help than night sleep |
The coffee nap
NIOSH endorses it explicitly: "You can gain the benefits of both caffeine and a short nap by consuming caffeine and then immediately taking a short nap." The rationale is the roughly 30-minute onset — you sleep through the lag, and the caffeine arrives as you wake.
NIOSH also notes that a brief nap of about 30 minutes can increase alertness during a night shift, while warning that sleeping much past 20 minutes raises the risk of waking groggy. It puts sleep inertia at typically 30 to 60 minutes, observed up to two hours, and cites research finding that 100 mg of caffeine on waking shortened it.
The underlying study is Reyner and Horne, 1997, in Psychophysiology. Twelve participants. Driving simulator. 200 mg of caffeine plus a nap under 15 minutes reduced driving incidents to 9% of the placebo condition, against 34% for caffeine alone.
That is a real result and NIOSH's endorsement carries weight, but twelve people in a simulator is a small foundation for a widely repeated claim. Treat it as a sensible, well-supported tactic rather than a proven law.
When to stop
NIOSH's shift-work training for nurses gives the warning without giving a number: "If you drink a cup of coffee near the end of a night shift, there can still be enough caffeine in your system to cause restlessness or waking while you are trying to sleep the morning after your shift." Elsewhere it advises avoiding caffeine "for several hours, according to individual sensitivity."
No fixed hour. That is deliberate, and given the two-to-ten-hour half-life range, it is the correct call.
What the research adds is a warning about your own judgment. A 2013 study in the Journal of Clinical Sleep Medicine gave twelve participants 400 mg of caffeine at 0, 3 or 6 hours before bed, measured with a validated in-home sleep monitor. All three timings significantly disrupted sleep versus placebo. At six hours before bed, total sleep time dropped by more than an hour — and the participants did not notice.
Two caveats you should carry with that finding. The sample was twelve people. And 400 mg in a single dose is roughly four times a single 8 oz cup — this is not evidence that one afternoon coffee costs you an hour of sleep. What it does suggest is that "I slept fine" is not reliable self-report.
Finding your own number
Move your last caffeinated cup 60 minutes earlier each week. Track how long you take to fall asleep and how often you wake. Give each setting a full week before judging it. Within a month you will have a personal cutoff that is worth more than any published guideline.
How much, per day
The FDA cites 400 mg a day as "an amount not generally associated with negative effects" for healthy adults — about two to three 12 oz cups of coffee. It notes wide individual variation, and it estimates toxic effects such as seizures around 1,200 mg consumed rapidly.
Worth stating clearly: the FDA does not publish a single-dose limit for healthy adults, and the 400 mg figure is a daily total, not a per-cup ceiling. It also does not apply universally — pregnancy, medications and existing conditions all change the picture, which is a clinician's territory rather than ours.
Honest limits
The American Academy of Sleep Medicine's 2025 guideline for shift work disorder suggests caffeine only as a conditional recommendation based on very low certainty evidence, and specifically cautions that caffeine near bedtime can worsen shift work disorder. Caffeine is a tool with a modest, uncertain evidence base — not a treatment.
Withdrawal is real and recognized in the DSM-5. Headache affects around half of people who stop abruptly, typically beginning 12 to 24 hours after the last dose, peaking between 20 and 51 hours, and lasting two to nine days. It can occur at habitual intakes as low as 100 mg a day. Taper rather than quit cold, and do it on days off.
And the sentence that governs the whole article, from NIOSH: caffeine "does not provide the alertness and decision-making ability that getting enough sleep provides." If your sleep is persistently disrupted, talk to a clinician. Shift work disorder and sleep apnea are both real, common and treatable, and neither responds to better coffee.
What to drink, given all that
Since roast level does not meaningfully change caffeine, choose on flavor and on when in the shift you are drinking.
Early, when you want brightness to wake your palate along with everything else: Last Call, our Ethiopian, Mexican and Brazilian light roast, all citrus and fruit. Mid-shift, when you want something balanced you can drink cup after cup: Grave Shift, the four-country signature medium — toffee, citrus, dark chocolate. For a heavier cup with real depth: Wake The Dead, our Colombian single origin, cherry and plum over chocolate.
And for the tail of the shift, when the ritual matters more than the stimulant: Final Breath, an Indonesian decaf with dark chocolate and full body. It contains a little caffeine, not none — the numbers are here.
To make any of them taste like the price you paid, use a ratio rather than a scoop. Our coffee-to-water calculator and guide to brewing a stronger, better cup cover the rest. If you would rather drink it cold on a Florida night, start with the cold brew guide.
The rest of the cluster
The overview is the best coffee for staying awake on long days and late shifts. Role-specific versions: nurses on twelve-hour shifts, truck drivers on long hauls, and permanent third shift.
Common questions
When should I stop drinking coffee before sleeping after a night shift?
There is no single correct hour. NIOSH advises avoiding caffeine "for several hours, according to individual sensitivity" and deliberately gives no number. Given an average half-life of about five hours, a reasonable starting point is to make your last caffeinated cup around the midpoint of your shift, then adjust based on how you actually sleep.
How long does caffeine stay in your system?
Half-life averages roughly five hours in healthy adults, but individual half-lives range from under two hours to about ten. That means a 200 mg dose could leave anywhere from a trace to a substantial amount in you ten hours later, depending on who you are. Oral contraceptives roughly double it; smoking shortens it by up to half.
Does a coffee nap work?
NIOSH recommends it: drink caffeine, then immediately take a short nap, so the roughly 30-minute onset coincides with waking. The main supporting study — Reyner and Horne, 1997 — found a large effect but had only twelve participants in a driving simulator, so it is a well-endorsed tactic rather than a settled finding.
Does caffeine six hours before bed really cost an hour of sleep?
That figure comes from a 2013 Journal of Clinical Sleep Medicine study in which twelve participants took 400 mg — roughly four times a single 8 oz cup — six hours before bed and lost more than an hour of total sleep without noticing. It is a real result, but the dose was large and the sample small, so it should not be read as one afternoon coffee costing an hour.
Does dark roast have more caffeine than light roast?
No, and the reverse claim is equally wrong as a flat statement. Brewed by weight, the 2024 Scientific Reports study found the whole light-to-dark spread within roughly 134–165 mg per 8 oz cup. Brew method, dose and grind matter far more than roast level.
Is more caffeine better when you are badly sleep-deprived?
The evidence says no. An Institute of Medicine review found 150 mg was about as effective as 300 or 600 mg in sleep-deprived subjects, and that caffeine restored alertness to only about 50% of the well-rested condition. Escalating the dose mostly buys side effects.
Better coffee for the hours you keep
Bean Reaper is a Florida brand that roasts to order through a U.S. roasting partner — your bag is roasted after you place the order and ships within two business days, which is why it tastes like something when it arrives. Twelve-ounce bags at $21.99, whole bean or ground, with a 30-day satisfaction guarantee.
Take a look at the full lineup when you have a free minute. The Reaper comes for stale coffee, not for you.