When to Stop Drinking Coffee for Better Sleep, According to Research

When to stop drinking coffee is a question I thought I had already settled. I have written before about my worst coffee habit, which is drinking it the moment I wake up. That one I am still working on.

The afternoon side of it I thought I had handled. My last cup is usually after lunch, and on genuinely busy days it stretches to four in the afternoon at the latest. That felt reasonable to me. Nowhere near bedtime, no obvious effect on falling asleep.

Then I found the study that made me reconsider whether “no obvious effect” means anything at all.

When to Stop Drinking Coffee: What the Study Found

A controlled study by Drake and colleagues, published in the Journal of Clinical Sleep Medicine in 2013, did something surprisingly simple that nobody had done before. Participants took a fixed dose of caffeine at three different times, right at bedtime, three hours before bed, and six hours before bed, with placebo on other nights. Sleep was tracked both by diary and by an in-home sleep monitor.

All three timings significantly disrupted sleep compared to placebo. That includes the six hour condition. Objectively measured total sleep time was reduced by more than an hour when caffeine was taken six hours before bed.

But the detail that actually stayed with me is this. The lead author noted that people are less likely to detect the disruptive effects of caffeine when it is taken in the afternoon. The monitors picked up the disruption. The participants largely did not.

That is the uncomfortable part. My four o’clock coffee on a busy day may well be costing me sleep I have no conscious awareness of losing.

The Honest Caveats

I want to be straight about the limits of this study, because it gets quoted a lot without them.

It included 12 healthy sleepers, which is a small sample. The dose was 400 milligrams taken all at once, which is roughly two to three cups of coffee rather than one. So the finding is not “one cup at four in the afternoon costs you an hour of sleep.” It is closer to “a substantial caffeine dose six hours before bed measurably disrupts sleep, and you probably will not notice.”

This is also why when to stop drinking coffee is not one fixed clock time for everyone. Individual variation also matters here more than in most areas of sleep research. Caffeine is metabolised at genuinely different rates between people, largely due to variation in a liver enzyme. Some people clear it quickly. Others take far longer, and those are the people for whom an afternoon coffee has an outsized effect.

Why Six Hours Is the Number

The reason when to stop drinking coffee has a specific answer comes down to arithmetic. Caffeine has a half life of roughly five to six hours in most adults. That means if you drink a coffee containing 200 milligrams of caffeine at four in the afternoon, around 100 milligrams is still circulating at ten in the evening.

Half of a coffee, still active, at the exact point your body is trying to wind down. Caffeine works by blocking adenosine receptors, and adenosine is the compound that builds up through the day to create sleep pressure. Blocking it does not remove the tiredness. It just stops you feeling it, while the pressure keeps accumulating underneath. That mechanism is the real reason when to stop drinking coffee matters more than how much you drink.

This is also part of why the cycle self-sustains, and why when to stop drinking coffee is worth getting right. Poor sleep leads to more caffeine the next day, which leads to more disrupted sleep. I went into the fatigue side of this in more detail in why you might always feel tired.

Healthier Afternoon Options That Do Not Wreck Your Sleep

Green tea contains meaningfully less caffeine than coffee, roughly a third as much per cup, and also contains L-theanine, an amino acid with a modest evidence base for supporting a calmer alertness rather than a jittery one. It is what I reach for in the late afternoon when I want something warm and mildly stimulating without pushing into the cutoff window. A loose leaf sencha keeps better than bags and tastes considerably less bitter if you brew it below boiling.

Decaf coffee is genuinely underrated here. It is not caffeine free, containing a small residual amount, but the difference is substantial enough that it works as a ritual replacement. If the appeal of the four o’clock coffee is the break and the warm cup rather than the stimulant, decaf gives you the whole experience without the cost. Worth choosing a Swiss Water process decaf over the cheapest option, since the solvent-based methods tend to taste noticeably flatter. It is the easiest swap once you have decided when to stop drinking coffee for the day.

Herbal tea is the cleanest option if you want something in the evening, since chamomile, rooibos and peppermint are all completely caffeine free. Not something I drink myself, but it is the most common swap people make once the cutoff moves earlier.

Sparkling water with lemon or a small amount of juice sounds unglamorous but works better than expected for the specific afternoon dip that is really about dehydration and monotony rather than tiredness.

Something to eat instead. This is the one I underestimated for a long time. A lot of the mid afternoon crash is blood sugar rather than a caffeine deficit, and a snack with protein and fat handles it more effectively than another coffee. I wrote about the specific options in low blood sugar snacks for afternoon fatigue.

What I Actually Changed

Not much, honestly. My own answer to when to stop drinking coffee was already close to reasonable.

What changed is that four in the afternoon is no longer my flexible upper limit on busy days. It is now the outer edge I try to stay inside of rather than push toward, and if I go to bed around ten, a two o’clock cutoff gives me the eight hour buffer that comfortably clears the six hour threshold. Treating when to stop drinking coffee as a boundary rather than a target was the whole shift.

The morning habit is still the one I have not solved. I wrote about why that one matters for cortisol rhythm, and it remains a work in progress.

Frequently Asked Questions

What time should I stop drinking coffee?

On when to stop drinking coffee, the research-supported minimum is six hours before bed. If you go to bed at ten in the evening, that means four in the afternoon at the latest, and earlier gives you more margin. If you are sensitive to caffeine or already sleeping badly, eight hours is a more cautious starting point.

Does one cup in the afternoon really matter?

It depends heavily on you. The study used a dose equivalent to two or three cups, so a single cup is a smaller input. But because caffeine metabolism varies so much between individuals, some people genuinely are affected by one afternoon coffee and others are not. The only way to know is to shift your cutoff for two weeks and see. If you are unsure, err earlier when deciding when to stop drinking coffee.

Why do I fall asleep fine after an afternoon coffee?

Falling asleep and sleeping well are different things. Caffeine can reduce deep sleep and total sleep time without making sleep onset noticeably harder, which is exactly why the participants in the study did not detect the disruption their monitors recorded. It is exactly why when to stop drinking coffee should not be judged by how quickly you drop off.

Is decaf coffee actually caffeine free?

No, it contains a small residual amount, typically a few milligrams per cup compared to around 95 to 200 in regular coffee. For most people that difference is large enough that decaf works fine in the evening.

Does tea count as coffee for the cutoff?

It counts, just less. Black tea contains roughly half the caffeine of coffee and green tea around a third. The same six hour principle applies, the dose is simply smaller.

To sum it up

Six hours before bed is the research-supported minimum for stopping caffeine. The study behind that number is small, and the dose was larger than a single cup, so it is a guideline rather than a law. If you take one thing from this, let it be that when to stop drinking coffee is a decision worth making deliberately rather than by habit.

The part worth taking seriously is not the number. It is that the participants could not feel the disruption their sleep monitors were recording. Falling asleep easily after an afternoon coffee does not mean it had no effect.

My cutoff was already close to the line. It is now a limit rather than a maximum.

What time is your last coffee, and have you ever tried moving it earlier? I would genuinely like to know.