Do you need to wait 90 minutes before morning coffee?
Where the 90-minute coffee rule came from, what the research shows, and how late is too late for your last cup.
The advice is to hold off on caffeine for 90 minutes to 2 hours after waking. People who follow it say that coffee too early gets in the way of your body waking itself up, and leads to a crash in the afternoon.
Andrew Huberman has recommended it on his podcast 1. An older version came from Steven Miller, then a neuroscience graduate student at the Uniformed Services University, whose 2013 blog post, later written up by his university, said coffee is most effective between about 9:30 a.m. and 11:30 a.m., after the morning cortisol peak 2.
Does coffee interfere with cortisol in the morning?
Cortisol rises sharply in the 30 to 45 minutes after waking. This is called the cortisol awakening response. In most healthy people, levels rise by roughly 40% to 75% above the waking value, then settle back down 34. What it is for is still debated 4.
The claim is that caffeine during this rise is wasted, or that taking it on top of high cortisol makes you build up a tolerance. Caffeine does raise cortisol. In a study by Lovallo and colleagues, people who had abstained from caffeine and then took it saw a clear cortisol increase, and in daily coffee drinkers that response was smaller but still present 5.
There's no published trial we could find that gives people coffee at waking versus 90 minutes later and measures alertness, the afternoon slump, tolerance or sleep. Miller's post was his own reasoning from how drug effects vary with time of day rather than an experiment, and the cortisol peak times it used are population averages that shift with each person's wake time 2.
Does waiting let adenosine clear first?
Another version of the rule says adenosine, the molecule that builds sleep pressure while you are awake, is still being cleared for a while after you wake up. Caffeine works by blocking adenosine receptors. The idea is that if you block them before the leftover adenosine is gone, it all hits you when the caffeine wears off, and that's the afternoon crash.
Adenosine does accumulate in the brain during long wakefulness and falls during sleep, shown most directly in animal studies 6, and caffeine's alerting effect comes from blocking its receptors 7. But most adenosine is cleared while you sleep, and there are no human data showing adenosine still clearing for 90 minutes after waking. There's also no routine way to measure adenosine in a living person's brain, so nobody has seen what it does in the first hours after waking.
Most people also have a natural dip in alertness in the early afternoon, coffee or not. Caffeine wearing off can make that dip feel sharper if the two happen around the same time.
Should you wait 90 minutes for coffee?
Waiting is fine if you like it, but no trial shows it prevents the afternoon slump.
If you have coffee as soon as you're up and feel fine through the afternoon, there's no good reason to change. Morning grogginess is more likely to improve with a steady wake time and bright light early on, covered in how much morning sunlight you need.
How long caffeine stays in your body
Caffeine's half-life in a typical healthy adult is around 4 to 5 hours, but the range is wide, from under 2 hours to about 9 in healthy adults 8. Most caffeine is broken down in the liver by the enzyme CYP1A2, and its activity varies with genetics. Smokers clear it much faster, and oral contraceptives slow it down. One study summarized in a 2018 review found an average half-life of about 7.9 hours in women taking them, against 5.4 hours in women who weren't. Pregnancy can push it well past that range, especially later in pregnancy 8. With an 8-hour half-life, half of a 4 p.m. coffee is still in you at midnight.
One often-cited test of late coffee is a 2013 study by Drake and colleagues. Twelve healthy adults who slept normally took 400 mg of caffeine, roughly the amount in four ordinary cups of brewed coffee, or a placebo at bedtime, 3 hours before bed or 6 hours before bed, and slept at home with a sleep monitor 9. Caffeine cut total sleep at all three times. Even at 6 hours, the monitor recorded more than an hour less sleep than placebo. The participants' own sleep diaries picked up the disruption when caffeine came at bedtime or 3 hours before, but at 6 hours they didn't report a significant difference 9.
A 2023 meta-analysis of 24 studies found caffeine reduced total sleep time by an average of 45 minutes and lowered deep sleep, and estimated that a standard coffee of about 107 mg should come at least 8.8 hours before bed to avoid losing sleep time 10. A 2025 trial in 23 men in their twenties found 100 mg made no measurable difference even 4 hours before bed. At 400 mg, falling asleep took longer when the dose came anywhere within 12 hours of bed, and sleep was more broken up within 8 hours 11.
What time to stop drinking coffee
Count back from your usual bedtime: about 9 hours for a regular coffee, and longer for a large cold brew or an energy drink. Tea, cola, chocolate, some painkillers and pre-workout powders all count toward the total. If you take hormonal contraception, are pregnant, or know caffeine keeps you up, set the cutoff earlier.
Pick a fixed cutoff time and stick to it, since you may not notice the lost sleep.
If you use Calibrate, your plan figures out this time from your bedtime.
When to see a doctor
If you rely on large amounts of caffeine to get through each day, feel your heart racing or get anxious after it, or have trouble sleeping that doesn't improve after you stop having caffeine later in the day, talk with a doctor. The same applies if you are pregnant or take medication that interacts with caffeine.
Turn this into a plan for your day.
Answer a few questions about your schedule and light access, and Calibrate builds a daily light-timing plan around the day you actually live.
References
- Huberman Lab (2022). Using caffeine to optimize mental and physical performance (podcast episode). www.hubermanlab.com/episode/using-caffeine-to-optimize-mental-and-physical-performance
- Uniformed Services University News (2017). The best time for a "coffee break" is not first thing in the morning. news.usuhs.edu/2017/09/the-best-time-for-coffee-break-is-not.html
- Pruessner JC, Wolf OT, Hellhammer DH, et al. (1997). Free cortisol levels after awakening: a reliable biological marker for the assessment of adrenocortical activity. Life Sci 61(26):2539-2549. doi.org/10.1016/S0024-3205(97)01008-4
- Stalder T, Kirschbaum C, Kudielka BM, et al. (2016). Assessment of the cortisol awakening response: expert consensus guidelines. Psychoneuroendocrinology 63:414-432. doi.org/10.1016/j.psyneuen.2015.10.010
- Lovallo WR, Whitsett TL, al'Absi M, Sung BH, Vincent AS, Wilson MF (2005). Caffeine stimulation of cortisol secretion across the waking hours in relation to caffeine intake levels. Psychosom Med 67(5):734-739. doi.org/10.1097/01.psy.0000181270.20036.06
- Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjørkum AA, Greene RW, McCarley RW (1997). Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 276(5316):1265-1268. doi.org/10.1126/science.276.5316.1265
- Fredholm BB, Bättig K, Holmén J, Nehlig A, Zvartau EE (1999). Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev 51(1):83-133.
- Nehlig A (2018). Interindividual differences in caffeine metabolism and factors driving caffeine consumption. Pharmacol Rev 70(2):384-411. doi.org/10.1124/pr.117.014407
- Drake C, Roehrs T, Shambroom J, Roth T (2013). Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed. J Clin Sleep Med 9(11):1195-1200. doi.org/10.5664/jcsm.3170
- Gardiner C, Weakley J, Burke LM, et al. (2023). The effect of caffeine on subsequent sleep: a systematic review and meta-analysis. Sleep Med Rev 69:101764. doi.org/10.1016/j.smrv.2023.101764
- Gardiner CL, Weakley J, Burke LM, et al. (2025). Dose and timing effects of caffeine on subsequent sleep: a randomized clinical crossover trial. Sleep 48(4):zsae230. doi.org/10.1093/sleep/zsae230