☕ Key takeaways
- Residual CO₂ works as a clock inside the bean. It empties fast over the first 24 to 48 hours, then slowly for weeks. Too much of it makes extraction erratic, none of it means the aromatics have already gone.
- On a medium roast the rest window runs 7 to 21 days for filter and 10 to 30 days for espresso. Light roasts start later and hold longer, dark roasts are ready sooner and stale faster.
- Three signals settle the verdict: the printed roast date, the size of the bloom on the first pour, and the smell of the beans when the bag is opened. A best-before date is not one of them.
Roast Freshness Guide: Degassing, Optimal Window, Storage
3 key takeaways
- Freshness has two edges, not one. A bag can be too young as easily as too old, and only one of those two problems fixes itself with time.
- The bloom is a free age test. A bed that rises and fizzes for 30 to 45 seconds still holds gas; a bed that quietly absorbs the water does not.
- Storage changes the slope of the loss, never its direction. Once the aromatics have evaporated or the oils have oxidised, no container gives them back.
Freshness is one of the most overlooked variables in home coffee making. Most people know to buy fresh bread, fresh milk, fresh vegetables, but coffee? That bag with a "best before 2027" date looks fine. The problem is that the best-before date on a coffee bag tells you almost nothing useful. What matters is the roast date, and this guide is built around a single practical question: is this bag still worth brewing, and what exactly is lost on either side of the window.
Rest window on a medium roast: 10 to 30 days for espresso, 7 to 21 days for filter, counted from the roast date. Past 45 days for espresso and 35 days for filter, the cup thins out noticeably. Store in an airtight, opaque container, dry and at room temperature. A best-before date says nothing about freshness, only the roast date does.
Why is residual carbon dioxide a reliable clock for coffee age?
Because what is left inside the bean falls away on a predictable curve. Wang and Lim measured roughly 6 mg of CO₂ per gram in light roasts, 11 mg in medium roasts and 15 to 16 mg in dark roasts on the day of roasting (Food Research International, 2014). The gas comes from Maillard reactions, caramelisation and pyrolysis during the roast, and stays trapped in the porous cell walls of the bean.
Emptying that reservoir happens in two stages: a fast release over the first 24 to 48 hours, then a long, slow tail lasting weeks. Three things set the pace. Storage temperature, since heat pushes gas out faster. Roast level, since darker beans are more porous and empty quicker despite holding more to begin with. And whether the coffee is whole bean or ground.
Grinding changes the timescale entirely. The same study recorded a loss of 26 to 59% of retained CO₂ at the moment of grinding, depending on how fine the setting was, with the remainder leaving within minutes. That is why pre-ground coffee should be judged by the hour it was ground, not by the date on the bag.
Which faults in the cup point to coffee that has not rested?
Inconsistency, mostly, which is what makes the fault easy to misread as a grinder problem. Below three to five days after roasting the bean is still saturated. Under espresso pressure the trapped gas forms a physical barrier between water and the soluble compounds: water slides around the bubbles instead of extracting evenly. The shot arrives under-extracted, short on sweetness and depth, and different every time. Crema looks generous but coarse, with large bubbles that collapse in seconds.
Pour-over shows the same excess from another angle. The first water onto very fresh grounds produces a fast, dramatic swell as CO₂ escapes under heat, and a bloom that violent carves preferential channels through the bed. Nothing is damaged, though. A few more days on the shelf fix a too-young coffee, which is never true at the other end of the window.
What exactly is lost once a coffee turns stale?
Acidity first, then sweetness, then the length of the finish, while a dry bitterness stays behind and takes up the space. A washed Ethiopian with no detectable acidity left has passed its window, whatever the label claims.
Two mechanisms produce that flattening, and they are worth separating. Oxidation is a reaction between oxygen and the lipid and aromatic compounds in coffee, turning them into molecules that read as rancid, cardboard, musty; it is irreversible. Volatilisation destroys nothing, it simply lets the lightest and most aromatic molecules evaporate, which is why coffee kept in an open jar goes quiet well before it smells rancid. Either way, no grinder, machine or recipe restores what has left.
How long should a light, medium or dark roast rest before you brew it?
Every figure below comes out of practice rather than out of a laboratory, so read them as tolerances. The medium roast column carries the reference windows used across this site; the other two columns apply one rule: a light roast holds less gas but releases it slowly, so it starts later and lasts longer, while a dark roast is ready sooner and stales faster.
| Method | Light roast | Medium roast | Dark roast |
|---|---|---|---|
| Espresso | 12 to 35 days post-roast | 10 to 30 days post-roast | 5 to 21 days post-roast |
| Pour-over (V60, Chemex) | 8 to 28 days post-roast | 7 to 21 days post-roast | 5 to 18 days post-roast |
| AeroPress | 7 to 30 days post-roast | 6 to 25 days post-roast | 5 to 20 days post-roast |
| Cold brew | 5 to 35 days post-roast | 5 to 30 days post-roast | 4 to 25 days post-roast |
| Moka pot | 7 to 35 days post-roast | 5 to 30 days post-roast | 4 to 21 days post-roast |
A well-sealed bag kept cool and dark sits at the generous end of each range, a bag opened daily in a warm kitchen sits at the tight end. The structural point is that espresso is the least forgiving method at the young edge and filter the least forgiving at the old edge.
What does a roast date prove, and what does it leave out?
It fixes a starting point and certifies nothing else. The date records the day the beans left the drum, not the temperature of the warehouse, not the weeks spent under a shop light, not whether the seal actually held. Two bags roasted on the same morning, one kept sealed at 15 °C and one sitting three weeks in a sunlit window, are not the same coffee. Here is what you will find on a label, from most to least useful:
- Exact roast date (e.g. "Roasted 08.04.2026"): ideal. You know precisely where you are in the window.
- Roast week (e.g. "Week 14 2026"): acceptable. Accurate to within a week, which is enough for filter and loose for espresso.
- Best-before date with the roast offset stated (e.g. "Best before 08.10.2026, roasted 6 months before"): acceptable if the offset is printed, useless otherwise.
- Best-before date alone: with no roast reference it carries no information. A bag reading "BBE 10/2026" could have been roasted in April 2025 or September 2026.
- No date at all: a sign of commodity coffee with no freshness accountability.
The best-before convention on industrial coffee is a commercial standard that allows months of warehouse storage without ever revealing the coffee's true age. Specialty roasters adopted the roast date as a transparency standard instead, which moves the burden of the verdict onto the buyer.
Is a valve on the packaging proof of freshness?
No. What a valve tells you is that the roaster sealed the bag early, nothing more. The small plastic circle on most specialty bags is a one-way degassing valve, patented by the Italian packaging maker Goglio in 1968. Its job is mechanical and dual:
- Lets CO₂ out: after packing, coffee keeps degassing. Without the valve, internal pressure would swell and eventually burst a hermetically sealed bag.
- Keeps air out: the passage only works one way, so oxygen does not enter. The coffee degasses without oxidising.
What the valve buys a roaster is the ability to seal a bag straight out of the drum instead of leaving it open for 24 to 48 hours. What it does not buy anyone is a date, a guarantee, or protection once the bag has been opened. A valve bag clipped shut carelessly takes in air like any other, and a six-month-old coffee sold with a valve is still six months old. Read the valve as a sign that the packer knows the trade, then go and find the roast date.
What can storage rescue, and what is already gone?
Once the bag is open, the only thing still under your control is the rate of loss. No container, vacuum or freezer returns a volatilised aroma or undoes an oxidation. Once the bag is open, all that is being decided is the slope of the loss:
- Airtight container: stainless steel canister, glass jar with a gasket seal, or the original bag closed tightly with a proper clip. The goal is to minimise exposure to oxygen and humidity.
- Room temperature (15 to 22 °C): avoid the refrigerator for daily use, since condensation from repeated opening does more harm than ambient warmth. The freezer suits a long-term stock, in sealed portions, with no round trips.
- Away from light: ultraviolet light accelerates the oxidation of surface oils. Use an opaque container or a closed cupboard.
- Away from moisture and strong odours: coffee adsorbs surrounding smells, spices, bread bags, cleaning products. A dedicated shelf prevents the transfer.
- Whole bean until brewing: grind no more than 15 minutes before extraction. Grinding multiplies the exposed surface and shifts the timescale from weeks to minutes.
Can you date a coffee by watching it bloom?
Closely enough to act on, and the test costs nothing but attention. Pour two to three times the coffee weight in water, say 30 g onto 15 g of grounds, then watch the bed for 30 to 45 seconds. A dome that climbs, cracks and fizzes means plenty of CO₂ remains, so the coffee is young, sometimes too young. A steady, moderate rise puts you in the middle of the window. A bed that absorbs water without a bubble has an empty reservoir, and the cup almost always confirms it. The bloom therefore earns its place twice over: it evens out the extraction, and it dates the coffee better than the label does.
Freshness isn't a connoisseur luxury, it's the minimum condition for the roaster's work to reach your cup intact. An exceptional coffee roasted three months ago will always be inferior to an ordinary coffee roasted last week. The roast date is the first line to read on any bag.