Caffeine Timing and Cortisol: Keep the Coffee, Change the Clock

Coffee has picked up a bad reputation in the cortisol corner of the internet. Drink it too early and you’re “spiking” your stress hormone. Drink it at all and you’re wearing out your adrenals. Wait 90 minutes after waking, or you’ll crash by lunch.

Some of that is true. Most of it, in my view, is overstated. Caffeine does raise cortisol, and regular drinkers only partly adapt. But the best-documented cost of coffee for a man with a long workday isn’t the morning cortisol bump. It’s what a late cup does to the night.

So this isn’t an article about quitting. It’s about timing: what the caffeine and cortisol research shows, where the evidence is thin, and a plan that keeps the morning coffee and moves the last one.

Yes, caffeine raises cortisol

The most useful study here is a 2005 crossover trial by Lovallo and colleagues. They took 96 healthy young men and women through 5-day phases at 0, 300 or 600 mg of caffeine a day. Each phase ended with a test day: challenge doses of 250 mg of caffeine or placebo at 9am, 1pm and 6pm, with saliva cortisol tracked across the day.

After 5 days without caffeine, the challenge doses clearly raised cortisol through the day. So the basic claim holds: caffeine stimulates cortisol release, at least in people who aren’t used to it.

A later paper from the same group looked at cortisol responses to mental stress, exercise and meals after caffeine. Caffeine raised cortisol and added to the cortisol response to a mental stressor, and the effect was smaller in habitual users. That’s worth knowing if your habit is a coffee to push through a stressful afternoon: on the cortisol side, the caffeine and the stress stack.

Regular drinkers adapt, but not all the way

Here’s the detail that gets lost. In the 2005 trial, after 5 days at 300 or 600 mg a day, the first morning dose produced no cortisol response. The body had adapted. But cortisol rose again after the 1pm dose. The authors concluded that habitual intake reduces the cortisol response to caffeine but doesn’t eliminate it.

Three limits. The volunteers were young and healthy, not men in their 40s. The “regular” phases lasted 5 days, not years. And the study measured saliva cortisol only. It shows the bump exists. It doesn’t show the bump harms a healthy person, and that’s a separate question the sources behind this site don’t answer.

The morning cortisol rise and the 90-minute rule

Cortisol has a daily rhythm without any help from coffee. According to an NIH-hosted Endotext chapter, it’s lowest around midnight, climbs through the later hours of sleep and peaks in the early morning. On top of that sits a separate burst called the cortisol awakening response, or CAR.

What the CAR is

Expert consensus guidelines for researchers define the CAR as the marked rise in cortisol over the first 30 to 45 minutes after you wake. A review by Fries and colleagues describes it as a distinct part of the stress system’s activity, layered on top of the daily rhythm. It’s influenced by sex, health status, health behavior and perceived stress. A review by Clow and colleagues explores a link between the CAR and the end of sleep inertia, the groggy stretch right after waking.

The consensus guidelines also make clear that measuring the CAR properly takes a tightly controlled sampling protocol. Keep that in mind the next time someone reads a lot into a single morning saliva test.

Where the 90-minute rule falls short

The rule goes like this. Your cortisol is already surging after you wake, so caffeine on top of it is either wasted or piles stress hormone on stress hormone. Wait 90 minutes, let the morning rise run its course, and you’ll dodge the afternoon crash.

It’s a tidy story. The evidence for it is thin. The research notes behind this site don’t include a trial that compared coffee on waking with coffee 90 minutes later and measured afternoon energy, crashes or cortisol. The 90 minutes doesn’t come from the CAR research either. The consensus guidelines put the CAR in the first 30 to 45 minutes after waking, and nothing in the sources here marks 90 minutes as a meaningful line.

The closest trial we have cuts against the premise, at least for regular caffeine users. In Lovallo’s study, after 5 days of daily caffeine, the 9am dose produced no cortisol response. That wasn’t a dose taken on waking, so it doesn’t settle the question. But if you drink coffee every morning, the dose that still moved cortisol in that trial was the 1pm one, not the first.

Get the 14-Day Fade Reset

Five short emails over two weeks: light, caffeine timing, walking, a sleep window and a stress log. Habits, not pills.

Email address

Leave this empty

Send me the Reset

Get the 14-Day Fade Reset. Double opt-in; unsubscribe anytime. Privacy policy.

Sign-ups open soon.

The bigger lever: your sleep

This is the part with better evidence. In a randomized, placebo-controlled trial, Drake and colleagues gave healthy sleepers 400 mg of caffeine at bedtime, 3 hours before bed or 6 hours before bed. All three timings disrupted sleep compared with placebo. At 6 hours before bed, objectively measured total sleep time fell by more than an hour. The authors concluded that people should avoid substantial caffeine for at least 6 hours before bedtime.

Twelve people and one big dose is a small base, and sleep was measured with a home monitoring device rather than a full sleep lab. The wider picture points the same way. A 2023 systematic review and meta-analysis pooled studies of caffeine and sleep. It found that caffeine reduced total sleep time and sleep efficiency and made it take longer to fall asleep. The authors also suggest keeping coffee out of roughly the last 8.8 hours before bed. That figure comes from a summary of the paper, not a full read, so take it as provisional until it’s checked.

Why this is also a cortisol story

Short sleep feeds back into cortisol. In a lab study of healthy young men, Leproult and colleagues looked at the evening after partial or total sleep deprivation. Evening cortisol was 37% and 45% higher, and the evening wind-down in cortisol started at least an hour later. The partial-deprivation night in that study was drastically shortened, a much bigger cut than a late latte usually costs you. So don’t map those percentages onto your week.

The direction still matters. If you care about your cortisol rhythm, I think protecting your sleep is a better-supported move than timing your first cup around the morning peak.

How much is too much?

The FDA cites 400 mg a day, about four or five cups of coffee, as an amount not generally associated with negative effects in healthy adults. That’s a general ceiling, not a target. Notice that Drake’s trial used that whole-day amount in a single dose, which may be part of why its sleep results were so clear.

If you have a heart condition, take regular medication, or find your sleep falling apart well below that amount, ask your doctor what makes sense for you. The FDA figure is a general guide for healthy adults, not a personal limit.

A practical caffeine plan

Count back from your bedtime

Set your last caffeine 8 to 10 hours before you go to bed. That window is my judgment, not a tested rule. Drake’s trial still found lost sleep at 6 hours and didn’t test longer gaps, and the provisional meta-analysis figure above is close to 9 hours. For a 10:30pm bedtime, that puts the cut-off between about 12:30 and 2:30pm, which is why “before 2pm” works as the short version for most office schedules. If you work late or go to bed after midnight, count back from your real bedtime instead of the clock on the wall.

Keep the morning coffee

Have your first coffee when it suits you. For regular drinkers, the cortisol data above give no strong reason to fear the morning cup. If you like the 90-minute wait, fine. Just don’t let it drag the whole caffeine day later.

Count everything

Energy drinks count. Pre-workout counts. So do tea, cola and the “just one” espresso after a client dinner. Some over-the-counter headache pills contain caffeine too, so check the active ingredients on the box. Read labels for the milligrams per serving. If a blend hides the amount, count it as a real dose and keep it out of the late slot.

Pre-workout is the easy one to miss, because it often goes down at 6pm before the gym. If you train after work, look for a caffeine-free version or move the session earlier.

Taper, don’t slam the door

Plan for 3pm without a coffee

The afternoon dip will still show up, because part of it looks built into your body clock rather than caused by lunch. A short walk outside or a 10-minute nap are the options I’d use in that slot. The reasoning is in why you crash every afternoon. The free 14-Day Fade Reset moves the caffeine cut-off to 2pm on day 3 and adds an after-lunch walk on day 6. If you’re not sure caffeine is your weak spot, the Fade Score asks about it along with nine other questions.

When caffeine is covering for something else

Sometimes the coffee habit is a symptom, not the cause. If you need caffeine just to stay awake while driving or in meetings, or you’re tired even after a full night most nights, timing tweaks probably won’t fix it. Get checked. Loud snoring, or someone noticing that you stop breathing in your sleep, is a reason to ask a clinician about sleep apnea. The Endocrine Society urges people with tiredness, weakness or low mood to get a real diagnosis. Obstructive sleep apnea and depression are among its examples.

If you’re waking at 3am night after night, that has its own set of causes, covered in waking up at 3am.

Bottom line

Caffeine raises cortisol. Daily drinkers adapt partly. In the best trial we have, after 5 days of regular caffeine, an afternoon dose still raised cortisol when the morning dose didn’t. The 90-minute rule is a tidy idea with thin evidence behind it.

The best-supported change is the boring one. Keep the morning coffee. Stop 8 to 10 hours before bed, which for most office schedules means before 2pm. Count every source of caffeine, pre-workout included. And if you’re moving the cut-off a long way, move it in steps.