Getting started

Caffeine before a sleep study: the cutoff is earlier than you'd guess, and there's an honest reason not to cheat

Sleep centers usually ask you to skip caffeine from noon or the afternoon on, and some say a full day. What a 400 mg study found about caffeine six hours before bed, and how to get through the evening.

A crescent moon above a coffee cup on a dusky blue and amber background

If you’ve been booked for an overnight sleep study, the prep sheet probably has a line that makes a daily coffee drinker wince: no caffeine after a certain time. For some people that time is noon. It seems harsh until you see what the night is for.

The short version

A sleep study records how you sleep on an ordinary night. Caffeine can change that night, so centers ask you to stop early. The Sleep Foundation says to avoid caffeine from the afternoon on, and the clinic instructions we saw in search results range from noon to a full 24 hours. This is general background, not medical advice. The sheet your center gave you is the rule.

Why the cutoff is so early

Caffeine lingers. As we explain in our piece on caffeine’s half-life, roughly half of a dose is still in your system around five hours later, though the figure varies a lot from person to person. A 3 p.m. coffee is still partly there at bedtime.

That is the same reason general sleep advice says to stop caffeine in the afternoon even without a study. The study just raises the stakes, because the point of the night is a clean recording.

What a 400 mg study found

In 2013, researchers at the Henry Ford Sleep Disorders and Research Center published a study in the Journal of Clinical Sleep Medicine. They gave people 400 mg of caffeine at bedtime, three hours before bedtime, or six hours before bedtime, and compared each with a placebo. Sleep was tracked at home with a validated portable monitor, alongside self-reports.

All three timings had significant effects on sleep compared with placebo. The authors concluded that caffeine taken six hours before bed has important disruptive effects. The American Academy of Sleep Medicine’s summary of the paper says total sleep time was reduced by more than an hour.

A few caveats. 400 mg is a fairly large dose, and people differ in how they handle caffeine. The study was done at home, not in a lab, and it didn’t test whether a small cup at noon matters for a given person. Still, it explains the logic: a cutoff many hours before bed is not arbitrary.

A single made bed with white linen and a grey blanket in a quiet room at dusk, beside a wooden nightstand holding an empty glass and a closed blue book
The goal of the prep: an evening that looks like a normal one, minus the caffeine. AI-generated illustration · image policy

The cheating problem

It’s tempting to sneak one coffee, especially if you’re already tired and anxious about sleeping in a strange room. There are two honest reasons not to.

  • It can blur the result. If caffeine shortens or fragments your sleep on the study night, the recording may not show how you usually sleep. That could mean a repeat visit.
  • Tell them if you did. If you had a cup after the cutoff, say so. A technologist would rather note it than guess.

The Sleep Foundation’s guidance also says not to nap that day, and to talk to your doctor before changing any medication, rather than skipping doses on your own.

Getting through the evening

For a daily drinker, an early cutoff means the afternoon slump arrives without its usual fix, and withdrawal can start within about a day for some people. A few things to weigh:

  • Taper in the days before, unless told otherwise. Going from three cups to one over a week means less to miss. Our no-headache tapering guide has the steps.
  • Move your caffeine to the morning. If your cutoff is noon, have the cup at breakfast and plan the rest of the day around it.
  • Keep the warm mug. Part of the habit is the ritual. A caffeine-free herbal coffee, such as a roasted chicory blend or Pero, can fill it; our roundup of alternatives has options. Read the label, since some blends contain barley (not gluten-free) or nuts.
  • Check the whole list. Many sheets name coffee, tea, cola, chocolate and energy drinks. Our guide to caffeine in decaf covers the trace amounts that can matter.
  • Skip the nap. It’s usually discouraged on study day, so plan a walk instead.

Worried about sleeping in a lab at all? That worry is common, and it’s worth mentioning to the technologist.

What to ask the sleep center

  1. What time is my caffeine cutoff, and does it include decaf and chocolate?
  2. Can I have a caffeine-free herbal drink in the evening?
  3. Which of my medications should I take as usual?
  4. What should I do if I get a withdrawal headache that afternoon?

If you’re preparing for other tests, we’ve covered caffeine before a cardiac stress test and before surgery. For the sleep science itself, see caffeine and sleep architecture.

Sources & further reading

  1. Caffeine effects on sleep taken 0, 3, or 6 hours before going to bed (opens in a new tab) — Journal of Clinical Sleep Medicine, 2013 (Drake, Roehrs, Shambroom, Roth)
  2. Caffeine in late afternoon, early evening can disrupt sleep (opens in a new tab) — American Academy of Sleep Medicine
  3. How Does a Sleep Study Work? (opens in a new tab) — Sleep Foundation

Reader conversation (4)

We read every response. Selected reader notes below.

  1. Dana K. · Portland, OR

    My sheet said nothing after noon. I’m a three-cup person, so I moved everything to 7 to 10 a.m. The afternoon was rough but fine.

    Editor reply · Sam Reyes

    Front-loading is the trick that works for most people. Thanks for sharing the schedule.

  2. Tomasz W. · Chicago, IL

    Does the no-nap rule really matter? I’m always wrecked by 4 p.m.

    Editor reply · Sam Reyes

    Ask your center. The general advice is to skip naps so you’re sleepy at bedtime, but they know their protocol and your reason for the study.

  3. Priscilla A. · Manchester, UK

    Mine was a home test rather than a lab. Same caffeine rule, which surprised me.

  4. Ben L. · Austin, TX

    I slept terribly in the lab anyway, but the tech said that is very normal. Worth knowing before you go.