Rest roasted coffee at least 12 to 24 hours, and in most cases somewhere between two days and two weeks. That spread isn't sloppiness on anyone's part. It's what happens when eight sources measure different things and each publish one number. It's also why a bag roasted yesterday can taste worse than one roasted last week.
Key takeaways
- A day is the floor and two weeks is the ceiling, and the honest answer for most coffee sits somewhere in between.
- What leaves the bean is carbon dioxide, and how much there is to lose depends on how dark the roast went.
- The field names roast level as the deciding variable and then can't agree on which direction it runs.
- Espresso is where the wait actually shows up in the cup, because pressure makes trapped gas visible in a way a drip brewer doesn't.
- The instrument that settles it isn't a published number, it's a real roast date and two brews from the same bag a week apart.
The honest range runs from one day to two weeks
The lowest credible floor anyone publishes is 12 to 24 hours. That one comes from Sweet Maria's coffee library, which also calls roasted coffee fresh for about seven days after that. It's the only page in the field giving a floor rather than a range, and the most defensible number on offer.
Then the range opens up. One widely republished guide puts filter coffee at 5 to 14 days and espresso at 7 to 14 days. Another won't call a light roast ready for espresso until past day 14. That's a factor of five between the first and third results, and neither acknowledges the other exists.
Something else is worth knowing about that page of results: it's smaller than it looks. Nine URLs, eight sources. The third and ninth results are the same article, heading for heading, running under two unrelated brands. And we're a roaster publishing about coffee we also sell, so read this on the same terms as the rest of it.
None of that is a scandal. It's what happens when nobody runs the same test twice.
What is actually leaving the bean
The gas is carbon dioxide, and roasting is what makes it. It forms inside the bean as the roast develops, most of it past first crack, and it starts leaving the moment the beans hit the cooling tray. A good deal of it stays put for weeks.
Xiuju Wang and Loong-Tak Lim measured how much there is to lose, in a paper titled Effect of Roasting Conditions on Carbon Dioxide Degassing Behavior in Coffee. In their arabica samples, light roasts held 6.29 to 6.70 milligrams of CO2 per gram, medium roasts 11.04 to 11.51, and dark roasts between 15.36 and 15.62. A dark roast starts with roughly two and a half times the gas a light roast does.
The second number should change how you think about waiting. The dark-roasted beans they measured needed more than 800 hours, about 33 days, to release approximately 90 percent of their trapped CO2. Degassing doesn't stop when your wait is over. It's still running while you're drinking the bag.
So being fully degassed isn't the target, and it can't be. What you're waiting for is the window where there's little enough gas left to brew cleanly and enough aroma left to be worth drinking. That's a judgment about a cup, not a countdown to zero. It's a different clock from the one green coffee runs on in the months it spends unroasted.
Everyone names roast level, and nobody agrees which way it runs
Three of the eight sources explain the spread the same way: roast level decides it. Read side by side, as they stood on 4 September 2026, the explanation comes apart. One says light roasts are ready in about 48 hours while dark roasts need up to a week. Another, on the same results list, says dark roasts are ready in one to five days and light roasts need 10 to 14 days or longer. Same variable, opposite sign, neither noticing the other.
A fourth page skips roast level and names something nobody else controls for: the machine. It puts fluid-bed roasts at 3 to 7 days and drum roasts at 8 to 14, which makes every day count published anywhere unportable to a roastery whose heat transfer differs. It's the most careful page in the group, and it doesn't notice what it just did to the rest of them.
The best evidence came from a roaster who ran the test instead of writing about it. Twenty-one coffees, roasted identically, cupped double blind against 24-hour controls on days 4, 8, 11, 17 and 22. About 73 percent peaked after eight days or more. One peaked at 24 hours. One was still improving on day 22. The individual coffee moved the answer further than any category did.
Here's where that leaves the most repeated rule in the field. The one real measurement of the gas says a dark roast has more of it to shed, not less. The rule that light roasts need longer is asserted by pages that measured nothing, contradicted by a page sitting next to them, and unsupported by the only laboratory figure available. Roast level does change a great deal about how a coffee tastes. Nobody has shown which way it moves the wait.
Espresso is where the wait shows up in the cup
Espresso is the one thing this field agrees on, and it's probably right. The pressure an espresso machine drives through a packed puck does something a drip brewer never does: it forces the trapped gas out all at once, in the cup, while the shot is running. You see it as a pale, blistered crema that collapses early, and you taste it as a shot that's sharp and hollow at the same time.
A drip brewer is far more forgiving. At atmospheric pressure the gas leaves upward while the bed is wet, and a young coffee that would wreck a shot will make a perfectly good pour-over several days sooner.
That's the honest hedge on our own shelf. Kenya Kiambu is marked for drip, not espresso, so if you're pulling shots it isn't the first bag on the lineup we'd point you to. And resting isn't always the problem worth solving. If the bag has no roast date printed on it at all, waiting won't fix anything, because the real question is how old that coffee already is.
Questions we get about resting coffee after roasting
How long should coffee rest after roasting?
Give it at least 12 to 24 hours, and expect most coffees to brew best between two and 14 days off the roaster. The floor is well established. The ceiling isn't, because it moves with the individual coffee more than with any category. If you can spare two days, take them.
Does espresso need a longer rest than filter coffee?
Yes, and it's the one point the whole field agrees on. Pressure forces trapped carbon dioxide out during the shot itself, so a young coffee gives pale, collapsing crema and a thin, sharp cup. Filter brewing at atmospheric pressure is much more forgiving. Published espresso windows run several days longer than filter windows.
Can you brew coffee the same day it was roasted?
You can, and it usually isn't worth it. Coffee brewed within hours of roasting tends to foam heavily, taste flat and finish short, because the carbon dioxide still coming out of the grounds gets between the water and the coffee. Sweet Maria's puts the minimum at 12 to 24 hours, the shortest credible wait anybody publishes.
Is resting the same thing as degassing?
Not quite. Degassing is what the coffee does, and it runs far longer than anyone waits. Resting is what you do, and it's finished when the cup is good rather than when the gas is gone. Wang and Lim measured dark-roasted beans still shedding gas past 800 hours, roughly 33 days, well beyond any sensible rest.
When does rested coffee stop improving and start going stale?
Most coffee turns the corner in the second or third week, and a bag you open every morning gets there sooner than a sealed one. Sweet Maria's calls roasted coffee fresh for about seven days. In one roaster's blind test a few coffees were still improving on day 22, so the corner isn't in the same place for every lot.
Start the clock on Wednesday
Every bag that leaves here comes off a Wednesday roast, which is why the date printed on it is an actual date and not a rounded guess, and it is also why the bag you open on Friday night is two days old, which is the half of freshness nobody warns you about. It's also why we can tell a wholesale account exactly which roast their coffee came off.
Kenya Kiambu is roasted light, and it's the bag on our shelf most likely to be judged on the wrong day, so read the roast date on it as a start line rather than an expiry. Brew it once on day two and once on day eight, same bag, same recipe, and decide for yourself. That comparison is worth more than any number published on this subject, ours included.
