You should stop drinking coffee 8 to 10 hours before your target bedtime if you are an average metabolizer, 12 to 14 hours before if you are a slow CYP1A2 metabolizer, and 5 to 6 hours before if you are a fast metabolizer. These numbers are derived from caffeine's 5-hour population half-life, adjusted for individual genotype, and calibrated against the 5–10 mg/L plasma threshold where sleep architecture begins to degrade. The "no coffee after 2 PM" rule is a statistical guess that applies to roughly one-third of the population.
I have been asked this question at least two hundred times. At dinner parties, in coffee shops, in DMs from strangers who found my website. Everyone wants a simple answer. The simple answer is wrong. The complex answer is: it depends on your liver, your bedtime, your dose, and what else is in your bloodstream. But people do not want a pharmacokinetic lecture. They want a time. So I built a calculator. And then I spent a year validating it against my own sleep data.
The math is not difficult, but it is unforgiving. Caffeine's half-life is 5 hours in the average healthy adult. After one half-life, 50% remains. After two, 25%. After three, 12.5%. After four, 6.25%. After five, 3.125%. Five half-lives is the standard pharmacological definition of "eliminated" — 97% clearance. At 5 hours per half-life, that is 25 hours total. But the sleep-relevant threshold is not zero. It is the plasma concentration at which adenosine receptor blockade measurably reduces deep sleep and REM. Polysomnography studies place this threshold at roughly 5–10 mg/L.
For a standard 200 mg dose in a 70-kg adult with average metabolism, the peak plasma concentration is approximately 8–10 mg/L. After one half-life (5 hours), it is 4–5 mg/L — below the sleep threshold. So an average metabolizer who consumes 200 mg at 8 AM will fall below the sleep-disruption threshold by 1 PM. They could drink another 200 mg at 2 PM and still clear below threshold by 7 PM, well before an 11 PM bedtime. The "no coffee after 2 PM" rule is actually conservative for average metabolizers. It is not wrong. It is just unnecessarily restrictive for some and dangerously permissive for others.
Here is the modeled cutoff by dose and metabolizer status for an 11 PM bedtime:
| Dose | Fast Metabolizer Cutoff | Average Metabolizer Cutoff | Slow Metabolizer Cutoff |
|------|------------------------|---------------------------|------------------------|
| 100 mg | 6:00 PM | 3:00 PM | 11:00 AM |
| 200 mg | 5:30 PM | 2:30 PM | 10:30 AM |
| 300 mg | 5:00 PM | 2:00 PM | 10:00 AM |
| 400 mg | 4:30 PM | 1:30 PM | 9:30 AM |
Look at the slow metabolizer column. At 400 mg — two large coffees — the cutoff is 9:30 AM. That is not "no afternoon coffee." That is "no coffee after your morning commute." I am a slow metabolizer. I learned this the hard way. For years, I drank a second cup at 10 AM and wondered why I could not fall asleep before 1 AM. My plasma caffeine at 11 PM was still 15 mg/L. I was pharmacologically awake. My brain was not receiving the sleep signal. And I blamed stress, blue light, my mattress — everything except the drug I had consumed twelve hours earlier.
What most people miss is that the cutoff is dose-dependent, not just time-dependent. The table above assumes a single dose. If you consume 100 mg at 8 AM and another 100 mg at 12 PM, your plasma concentration at 11 PM is the sum of the residual from both doses. The first dose has undergone three half-lives (12.5% remaining = 12.5 mg). The second dose has undergone two half-lives (25% remaining = 25 mg). Your total residual is 37.5 mg. In a slow metabolizer, that is still above the sleep threshold. The cumulative effect of multiple doses pushes the effective cutoff earlier than any single-dose calculation would suggest.
This is why I recommend a caffeine cutoff calculator rather than a simple rule. The calculator models multiple doses across the day, accounts for your metabolizer status, and plots your plasma concentration curve against your bedtime. It tells you not just when to stop, but what your sleep risk is if you have an afternoon espresso. The difference between "probably fine" and "definitely not fine" is sometimes 50 mg and 2 hours.
The medication factor is also underappreciated. Oral contraceptives increase caffeine's half-life by 30–70% in most women. Fluvoxamine can increase it by 500%. Pregnancy triples it. If you are on any of these and using the average metabolizer cutoff, you are sabotaging your sleep. A woman on birth control who drinks coffee at 2 PM is metabolizing it like a slow metabolizer. Her effective cutoff is 10 AM, not 2 PM. She follows the "no coffee after 2 PM" rule and still lies awake at midnight, wondering what is wrong with her.
Sam — fast metabolizer, no medications — drinks tea at 9 PM and sleeps fine. I — slow metabolizer, no medications — drink coffee at 10 AM and that is my last dose. We live in the same house. We go to bed at the same time. Our cutoff times are 12 hours apart. This is not a lifestyle difference. It is a biochemical difference. And pretending otherwise is how people develop chronic sleep debt, anxiety, and the mistaken belief that they are "just bad sleepers."
The practical answer is: know your metabolizer status, know your dose, and model your clearance. If you do not know your genotype, assume you are average — but understand that 55% of the population is slow, so "average" is a gamble. Start with a conservative cutoff: 10 hours before bed for your last dose. Track your sleep for two weeks. If you fall asleep easily and wake rested, you might be a fast metabolizer and can push later. If you lie awake, you are probably slow and need to pull earlier. The data will tell you. The math will confirm it.
🌙
Sleep Impact Estimator
Enter your bedtime, dose, and metabolizer status to get your exact cutoff time.
All data stays in your browser — we never see it.⏱️
Caffeine Half-Life Tracker
Model your plasma caffeine curve across the day and see your bedtime concentration.
All data stays in your browser — we never see it.
### Frequently Asked Questions
#### What time should I stop drinking coffee for sleep?
For an 11 PM bedtime: fast metabolizers can stop at 5:30 PM, average metabolizers at 2:30 PM, and slow metabolizers at 10:30 AM. These times assume a 200 mg dose and adjust proportionally for larger or smaller doses.
#### Does the cutoff time change based on how much coffee I drink?
Yes. A 100 mg dose clears faster than a 400 mg dose. The cutoff moves earlier by approximately 30 minutes for every additional 100 mg. Multiple doses across the day have cumulative effects that push the effective cutoff earlier than any single-dose calculation.
#### Can I drink decaf coffee in the evening?
Decaf contains 2–15 mg per cup. For fast metabolizers, this is negligible before bed. For slow metabolizers or caffeine-sensitive individuals, even 10 mg at 8 PM can produce measurable sleep disruption. If you are sensitive, switch to herbal tea after your cutoff time.
#### How do medications affect the coffee cutoff time?
Oral contraceptives extend half-life by 30–70%, effectively making you a slow metabolizer. Fluvoxamine can extend it by 500%. Pregnancy triples half-life. Any of these conditions requires cutting your caffeine cutoff 3–6 hours earlier than the standard recommendation.
#### Is "no coffee after 2 PM" a good rule?
It is conservative for average metabolizers and dangerously permissive for slow metabolizers. A better approach is to calculate your personal cutoff based on dose, metabolizer status, and medications. For slow metabolizers, the safe cutoff is often 10 AM or earlier.
---
From the loft, where my coffee cutoff was 10:15 AM and it is now 9:47 PM. I will be asleep by 11.
Caffeine is a tool, not a lifestyle. Measure it.
