How long roasted coffee actually stays good
Short answer: the date on the bag is a food-safety and labelling date, not a quality window. Roasted coffee stays safe far longer than it stays good, and it also needs time before it is at its best. Quality is set by three processes on different clocks — gas leaving, oxygen arriving, moisture arriving — and the only reliable way to find your own window is to taste one batch across its life.
What does the date on a coffee bag actually mean?
A "best before" date is a declaration made under food labelling rules that the product is expected to remain safe and wholesome up to that point. Roasted coffee is dry, low in water activity and sterilised by the roast itself, so that bar is easy to clear, and the date carries a margin and assumes an unopened package.
None of that describes flavour. A bag can be well inside its date and taste flat, or past it and still make an acceptable drink. Conflating the two produces both common errors: coffee used long after it stopped tasting of anything, and good coffee discarded because a printed date passed.
The roast date on many bags is directly relevant to quality, but it is only a reference point: it tells you where the coffee is on its clocks, not how long those clocks run. That depends on the coffee, the roast, the packaging and your storage.
What is actually changing inside the bag?
Three processes, running simultaneously at different rates.
Carbon dioxide leaves. Roasting generates a large quantity of CO2 inside the bean, and most of it does not escape at the end of the roast. It stays trapped in the cell structure and works its way out gradually. This is degassing, and it carries aroma with it — the smell that hits you when a fresh bag is opened is gas leaving, taking volatile aromatics along.
Oxygen gets in and reacts. Coffee contains lipids and a very large number of volatile aromatic compounds, and both react with oxygen. Oxidised lipids read as stale, cardboard, crayon, sometimes faintly rancid. Aromatics oxidise or simply evaporate, which is why an old coffee smells of less than a new one long before it tastes actively bad.
Moisture is taken up. Roasted coffee is drier than the air around it in most conditions, and it is porous, so it absorbs water vapour. That water carries oxidation further and faster and changes how the bean fractures in the grinder. Damp coffee also picks up whatever else is volatile nearby, which is why storing coffee beside strong-smelling stock is a bad idea.
These do not all push in the same direction. Degassing improves how coffee behaves in an espresso machine, for a period — a coffee full of gas is genuinely difficult to extract well. Oxidation and moisture uptake only ever make coffee worse, from the moment the roast ends. So the trajectory is not a decay curve from a peak on day one; it is one improving process overlapping two deteriorating ones. The window is where the first has done enough and the others have not yet done too much.
Why does fresh coffee need to rest — and why does espresso need more rest than filter?
Very fresh coffee releases CO2 during extraction, and in an espresso machine that is a problem. Gas coming out while water is being forced in obstructs saturation, so parts of the bed stay dry longer than others. Gas released under pressure expands the puck mid-shot, so bed resistance changes while the shot is running. And gas escaping upward finds or creates paths of least resistance, which water then follows — the mechanism set out in espresso channelling explained.
The result is a shot that is fast in places and slow in others, often sour because well-extracted and under-extracted fractions arrive in the same cup, and different every time even with identical preparation. Crema on very fresh coffee is large, foamy and quick to collapse, being mostly carbon dioxide rather than the emulsion and fine particulate that gives a stable crema its structure — see does crema still matter.
As gas leaves, saturation becomes even, bed resistance becomes predictable and the shot stabilises. Nothing about the recipe has changed; the coffee has stopped fighting the water. Rest is a recipe variable rather than a preference, and sits alongside the others in the complete guide to espresso extraction variables.
Filter brewing is far more forgiving: no pressure, long contact time, an open bed where gas escapes upward without disturbing much, and a bloom stage that exists precisely to let it go. Coffee that produces erratic espresso can make an excellent filter brew the same day. If you run both, expect them to want different rest from the same bag; if you only run filter, much of the espresso anxiety about freshness does not apply to you.
What moves the window?
The direction of each of these is predictable. The length is not, which is why the method further down matters more than any figure.
| Variable | Direction |
|---|---|
| Roast level | Darker roasts are more porous and open, so they release gas faster — shorter rest — and admit oxygen faster, so they stale sooner. Light roasts are dense, hold gas longer, need more rest, then hold up longer. |
| Bean density and origin | Denser, higher-grown coffees behave like light roasts here: slower to give up gas, slower to deteriorate. Softer, lower-density beans move faster in both directions. |
| Whole bean vs ground | Grinding accelerates everything enormously. See below. |
| Packaging | A one-way valve lets CO2 out without letting air in, which is why bagged coffee does not swell. Nitrogen flushing displaces oxygen before sealing; vacuum removes it. Both slow oxidation substantially while the package is intact, and neither does anything once it is open. |
| Storage conditions | Heat, light, air and humidity all accelerate oxidation and moisture uptake. Repeatedly opening one large bag is worse than working through small ones. |
Two special cases. A blend assembled after roasting, from components roasted separately to different levels, ages as several coffees at once — the darker components move ahead of the lighter ones, so the balance shifts as the bag gets older, not just the intensity. That is one of the trade-offs in how to roast a blend. Low-turnover sites face the problem from the other end: a machine serving a modest number of drinks a day holds the same coffee for a long time however it was bought, which is part of why roast choice there is its own question — see is there an ideal roast profile for super-automatic machines.
Why ground coffee is a completely different question
Grinding multiplies exposed surface area by an enormous factor, and everything above scales with surface area. Gas that would have taken days to work out of a whole bean is gone quickly, and aromatics held inside a closed structure meet air all at once.
The honest answer for ground coffee at peak quality is minutes to hours, not days. That is what the surface-area change implies, and it is audible in the cup if you grind two doses and brew one immediately and one an hour later. Pre-ground coffee in a sealed pack can be serviceable for a long time as a beverage; it will not be at peak quality after the first few minutes of air. See whole bean vs pre-ground: what happens after grinding.
How should coffee be stored?
Airtight, cool, dark, dry, and away from anything with a strong smell. Those five cover almost all of it, and the order matters less than doing all five.
The fridge is a bad idea. A fridge is humid, and coffee taken out of it and opened is colder than the air around it, so moisture condenses onto and into the beans. A fridge is also full of volatile compounds from other food, and porous coffee is an efficient way to collect them.
The freezer is a different argument and should not be lumped in with the fridge, because the mechanism is not the same — arresting the reactions rather than slowing them, with its own handling rules. That case is treated separately in freezing coffee: what the evidence says.
What should an operator actually do about it?
Order to your turnover, not to a delivery schedule. Buying to a convenient delivery rhythm rather than to what you actually sell is how bars end up holding more coffee than they can use while it is good. If your volume does not fit the delivery cycle, the cycle is the thing to renegotiate.
Date-mark on arrival as well as on roast. Roast date tells you where the coffee is on its clocks; arrival date tells you how long it has been in your conditions, which is the part you control. Both on the bag, in marker, takes seconds.
Rotate properly. Oldest bag forward, every delivery, no exceptions. The failure mode is undramatic: one bag left at the back while newer stock is worked through in front of it.
Keep the hopper at working quantity, not full. A full hopper is a large mass of coffee sitting warm, in light and exposed to air for as long as it takes to sell. Filling it is convenient and costs quality every time — see keeping coffee fresh in the hopper.
Re-dial as a bag ages, rather than assuming the grinder has drifted. As coffee degasses and then oxidises, its behaviour under water changes and the same setting gives a different shot. That is the coffee moving, not the burrs, and knowing the difference saves a lot of unnecessary adjustment — see can you dial in grind size by taste alone, while genuine mechanical drift belongs in grinder maintenance and calibration schedules.
How do you find your own window?
A method, because a number would be a guess.
Buy one batch of one coffee, large enough to last well beyond the point where you expect it to be good. Taste it on a schedule — a fixed interval, written into the diary, not when you happen to remember — in the drink you actually sell most of. If that is a cappuccino, taste a cappuccino; milk covers some faults and exposes others, so the window for a milk drink is not the window for a straight espresso.
Write down the date, the recipe, and two or three words about the cup, on paper on the bar rather than in someone's head. You are looking for two boundaries: where the shots stop being erratic, and where the flavour starts flattening out. The middle is your window, for that coffee, in your conditions, on your equipment.
Then repeat with a coffee at a different roast level, since that is the variable that moves the answer most. Two runs give you a usable rule for your bar, at the cost of one bag and some attention — and they replace the figure you were hoping to find in an article.
Where freshness obsession starts costing money
Freshness has become a proxy for quality, and past a certain point it is a poor one.
A bar throwing away coffee because a date passed is destroying product and money for no gain in the cup. If the coffee tastes good, it is good; if it does not, the date is not why. Discarding by calendar rather than by taste is worth examining, particularly at the volumes a busy site goes through.
"Roasted last Tuesday" on a menu is marketing, not a quality guarantee. It says something about logistics and nothing about the coffee, the roast, or the person making the drink. A well-chosen, well-roasted coffee three weeks old will comfortably beat a mediocre one three days old. Freshness is a necessary condition for a very good cup and nowhere near a sufficient one.
And most quality problems blamed on staleness are not staleness. They are grind — drifted, or never right for the coffee. They are dose, weighed once and assumed thereafter. They are water, untreated or running through an exhausted cartridge. Those three account for the great majority of "the coffee isn't tasting right" calls, and all three are cheaper to fix than a change of supplier. Check them before you blame the bag.
Frequently asked questions
Is coffee past its best-before date unsafe? No. It is a dry roasted product and the date is a labelling declaration with a margin, not a spoilage cliff. Taste it. If it tastes flat, that is a quality decision, not a safety one.
Should I let a new bag rest before putting it on the bar? For espresso, usually yes — very fresh coffee produces erratic shots for mechanical reasons rather than flavour ones. How long depends on the roast level and the coffee, which is what the tasting method above is for. For filter it matters much less.
Does a valve bag keep coffee fresh after opening? No. The valve only lets gas out; once the bag is open, air is in and the packaging is doing nothing for you. Reseal it tightly and buy sizes you will finish reasonably quickly.
Can I keep coffee in the fridge? Better not. Fridges are humid and full of other smells, and porous coffee collects both. Airtight, cool, dark and dry in a cupboard beats a fridge. Freezing is a separate question with a different mechanism.
My shots have changed and the bag is only part-used. What happened? Most likely the coffee, moving along its degassing and oxidation curves — and it will keep moving. Re-dial to the cup rather than assuming the grinder has drifted; if a fresh bag of the same coffee behaves the same way, then look at the equipment.
This article is part of the Vea Group knowledge base. Vea Group S.p.A. is an Italian designer and manufacturer of professional and premium coffee machines, with heritage dating to 1919 and production in Chignolo d'Isola, Italy, and Suzhou, China.