For most people, no.
A standard cup of decaf contains 2 to 15mg of caffeine. A cup of regular coffee runs 70 to 140mg, often more for espresso. The clinical research on caffeine and sleep tests doses of 100 to 400mg, not 5mg. The trace caffeine in decaf sits well below the dose shown to disrupt sleep in the average adult.
That is the headline. The detail underneath it is where the interesting bits live: who might still feel something, why the timing rules for regular coffee don’t apply, and which decaffeination method leaves the least behind.
How much caffeine is actually in decaf?
Between 2 and 15mg per 8oz cup. For context, a regular cup is 70 to 140mg. Decaf works out to roughly 10% of the caffeine in a caffeinated cup, often less.
The reason the range is so wide:
- Decaffeination method. Swiss Water Process removes 99.9% of the caffeine. Sugar cane and standard solvent methods remove around 97%. That difference is roughly 3mg per cup at the upper end.
- Bean type. Robusta beans have around twice the caffeine of Arabica before decaffeination, which leaves more residual behind.
- Roast and brew. Espresso extraction concentrates caffeine into a smaller volume. Light roasts retain slightly more caffeine than dark.
- Cup size. A 16oz cup of decaf averages around 9.7mg. An 8oz cup of Swiss Water processed decaf often comes in at 2 to 7mg.
The most cited primary study is University of Florida (2006), published in the Journal of Analytical Toxicology. It found espresso decaf shots contained 3.0 to 15.8mg per shot, and brewed decaf from various outlets ranged from 0 to 13.9mg per 16-oz serving. That variability has held up across subsequent testing.
FDA rules require at least 97% of the original caffeine to be removed. EU Directive 1999/4/EC uses a different standard, capping residual caffeine by weight rather than setting a percentage-removal requirement. Swiss Water typically clears 99.9% by mass.
Will decaf coffee keep you awake?
Not for most people. The amount of caffeine needed to disrupt sleep is much higher than what’s in a cup of decaf.
Gardiner et al. (2024), published in SLEEP, tested caffeine doses against placebo. 100mg produced no significant effect on sleep. 400mg, consumed within 12 hours of bedtime, delayed sleep onset and altered sleep architecture. The earlier Drake study (2013) behind the “stop coffee by 2pm” rule used 400mg, six hours before bed. That dose is 26 to 200 times what’s in a cup of decaf.
A 2017 randomised double-blind trial in the Journal of Caffeine Research (Distelberg et al.) is the cleanest test of decaf specifically. 49 adults drank either caffeinated or decaffeinated coffee for five days, in a 15-day crossover study. The caffeinated group reported impaired sleep and increased anxiety. The decaf group showed no significant change.
Caffeine works by blocking adenosine receptors in the brain. Two to five mg is not usually enough to block them noticeably.
Who might still feel the effects?
A small group. Three categories worth naming.
Slow metabolisers. A minority of people carry CYP1A2 gene variants that extend caffeine’s half-life from the usual 3 to 5 hours up to 8 to 12. Same dose, longer to clear. For this group, even a small evening dose can linger.
Receptor sensitive. The ADORA2A gene controls how sensitive your brain’s adenosine receptors are. Some variants produce a noticeable response at doses that would not register for most people.
Recently caffeine free. Tolerance drops fast. Someone who has been off caffeine for several weeks may feel 5mg as mild stimulation. The body has stopped buffering for it.
Pregnant women are worth a separate note. NHS guidance caps daily caffeine at 200mg during pregnancy. Decaf sits well within that limit even across multiple cups, and the NHS names decaf as an acceptable drink. If you want exposure as close to zero as practical, choose a Swiss Water processed decaf at around 2 to 5mg per cup.
When to drink decaf if sleep matters to you
The standard “stop caffeine 8 to 10 hours before bed” rule is calibrated for regular coffee, where a single cup delivers 100mg or more. It doesn’t apply to decaf in any meaningful way.
For the average decaf drinker, an evening cup is fine. A cup before bed is fine. The maths doesn’t bite. A 5mg dose at 9pm, with a half-life of roughly 5 hours, leaves about 1.25mg by 7am. That is well below the dose any sleep study has flagged.
If you are a slow metaboliser or particularly sensitive, choose a Swiss Water processed decaf and finish your cup an hour or two before bed. The warm drink itself has a small sleep-promoting effect via the body temperature route: temperature rises during drinking, then falls, which mirrors the natural pre-sleep cooling phase.
The practical case for decaf in the evening is the cup itself, not the caffeine. It preserves the coffee ritual without the stimulant load. For people who have cut their caffeine intake but miss the cup, that matters more than the milligrams.
Does the decaffeination method affect caffeine content?
Yes, though the differences are small in practice.
| Method | Caffeine removed | Typical per cup |
|---|---|---|
| Swiss Water Process | 99.9% | 2 to 5mg |
| CO2 (supercritical) | 99% or higher | 2 to 5mg |
| Sugar cane (ethyl acetate) | ~97% | 3 to 9mg |
| Standard solvent (MC) | ~97% | 3 to 9mg |
Swiss Water and CO2 are the two specialty grade options. Both leave roughly the same residual caffeine behind. Sugar cane, often called natural decaf because the ethyl acetate is derived from fermented sugarcane, removes slightly less. Standard solvent methods (methylene chloride or ethyl acetate from chemical synthesis) also clear around 97%.
For most people the gap between 3mg and 7mg changes nothing. For slow metabolisers, the recently abstinent, or anyone optimising for the lowest possible dose, look for “Swiss Water Process” on the bag. How Swiss Water works covers the mechanism in full.
Decaf vs caffeine free: what’s the difference?
Decaf coffee started with caffeine and had most of it removed. In the US, FDA rules require at least 97% removed; in practice 2 to 15mg per cup remains. Caffeine free means made from a plant that never contained caffeine in the first place. Rooibos, chamomile, peppermint, hibiscus, chicory root and dandelion root all contain zero milligrams.
Worth checking the label. Anything containing Camellia sinensis (green tea, black tea, white tea, oolong) has caffeine even in a “herbal” blend.
The distinction matters mainly in three situations. You have cut caffeine entirely and even 5mg feels like something. You are pregnant and prefer zero rather than under 200mg. You are highly caffeine sensitive.
For everyone else, the trace caffeine in decaf is irrelevant to sleep.
Our verdict
Decaf does not keep most people awake. It is safe in the evening, safe before bed for the average drinker, and the clinical evidence on this is consistent. The 2 to 15mg dose is not enough to block adenosine receptors meaningfully in a typical adult.
If you fall into the slow metaboliser, hyper sensitive or recently abstinent group, choose a Swiss Water or CO2 processed coffee and drink it an hour or two before bed rather than directly before. That covers the practical edge cases without overcomplicating things.
We have tasted decaf coffees from UK roasters across the directory. The ones we recommend specifically for evening drinking are the Swiss Water and CO2 processed lots. Browse the directory and filter by decaf method if you want the lowest caffeine options first.