Wine Cellar Conditions: Temperature, Humidity, Light

Temperature is the only cellar variable with enough published evidence behind it to justify serious money. Fix it, hold it steady, and the rest of the list is housekeeping.

Wine cellar conditions come down to five things: temperature, temperature stability, humidity, light and vibration. The Australian Wine Research Institute, citing Amon and Simpson, recommends bottled wine be stored "with the cork in contact with the wine" in a cool, dry location, and puts cool at 15 to 20°C. The Comité Champagne recommends 10 to 15°C, 60 to 80% humidity, a well ventilated place, and protection from light, vibrations and odours. The ceilings matter more than the targets. The AWRI warns that any storage place where the temperature "exceeds 25°C for long periods and 40°C for short periods can affect wine quality", and that above 40°C the visual and sensory changes arrive "in only a matter of days". Light moves faster than collectors expect: in Dozon and Noble's trial, a still white wine in green glass carried a significant lightstruck fault after 31.1 hours under fluorescent lamps, according to AWRI Technical Note TN09. Vibration is the only item on the list resting on a single study.

What are the ideal wine cellar conditions?

Here is the whole list, with the published number, the point at which each one starts to cost you, and who says so.

Condition	Target	Where it starts to hurt	Who publishes it
Temperature	10 to 15°C (Comité Champagne); 15 to 20°C (AWRI)	Above 25°C for long periods, above 40°C for short ones	AWRI, Comité Champagne
Stability	Constant, no day-to-night cycling	Thermal expansion moves corks and causes leakage	AWRI
Humidity	60 to 80% (Comité Champagne); 75% is the figure most often quoted	Dry enough to shrink a cork; damp enough to cost you the label	Comité Champagne, Oxford Companion
Light	Dark, or bottles kept in cartons and wooden cases	3.3 hours under fluorescent lamps in clear glass	AWRI Technical Note TN09
Vibration	Minimise	18 months at raised vibration accelerated changes in acids and tannins	Chung and colleagues, 2008
Position	Cork in contact with the wine	Little measured difference across 60 months	AWRI

Temperature and stability carry the money. The other four are cheap to get right.

Which condition actually costs you money?

Heat, then instability. Everything else is a rounding error by comparison.

The ranking follows the size of the published effects. The AWRI's technical note on heat and light cites Marais's 1986 work. Marais found faulty flavours and falling overall quality after twelve months' storage at 30°C. That is not a freak event. That is a warm room. The same note puts Ough's finding alongside it. Ough reports that above 40°C, visual and sensory changes arrive in days. Nothing in the light or vibration literature moves at that speed.

Light sits third because it is cheap to solve. Keep the cases shut and the problem disappears. Humidity sits fourth because the evidence for any particular percentage is thinner than the confidence with which people quote it. Vibration sits last, on one study.

The market prices this asymmetry, and it prices it through condition rather than through a storage certificate. A bottle that has been hot shows it: seepage under the capsule, a pushed cork, a fill level that runs ahead of the bottle's age. That is why ullage bands do the work at auction that a storage claim cannot.

What temperature should a wine cellar be?

Somewhere between 10 and 15°C, held constant. That is where the published recommendations cluster once you set the outliers aside.

Three standard references land in the same place, as collected's survey of wine storage. Jancis Robinson's Oxford Companion to Wine recommends constant temperatures between 10 and 15°C. Tom Stevenson states that the ideal is 11°C. Karen MacNeil's Wine Bible says 13°C. The AWRI sits higher at 15 to 20°C, which answers a different question: it describes acceptable storage rather than optimal ageing.

The mechanism behind all of it is reaction rate. Venturi and colleagues, writing in 2019, set wine ageing out as an Arrhenius problem: as temperature rises, so does the rate of the reactions. The twelve-month trial found the ageing of a red wine "significantly delayed" at 4°C against 20°C, across every packaging type but one, according to Venturi and colleagues. The exception was the wine in glass bottles under natural cork, which is worth knowing before you read the result as settled. a useful rule of thumb is that the rate of chemical reactions in wine doubles for every 10°C.

So the cost of a warm cellar is not a fault. It is time. A cellar at 20°C runs your wines at roughly twice the clock speed of one at 10°C, and every drink window you have read was written for a bottle kept cool.

Two pieces of reassurance sit at the edges. Wine can be stored as warm as 21°C without long-term harm, and Professor Cornelius Ough of UC Davis held that a few hours at 49°C will not damage a wine. A hot afternoon in transit is survivable. A hot summer is not. The full set of numbers, including what to set a cabinet to, sits in the wine storage temperature guide.

Do temperature swings matter more than heat?

For the bottle's seal, yes. Constant warmth ages wine faster. A swinging temperature physically works the cork.

The AWRI states the mechanism plainly: "leakage of wine and/or movement of cork stoppers due to thermal expansion of wine" follows exposure to temperatures significantly greater than ambient, and "thermal cycling where the temperature varies significantly between day and night should be avoided". Wine expands, the headspace pressurises, and the only moving part in the system is the cork.

There is now a measurement of where that seal fails. Chanut and colleagues, publishing in PNAS Nexus in 2023, tracked oxygen entering bottles over 24 months. With model wine in contact with the stopper, the glass-to-cork interface, not the cork body, carried nearly 75% of total oxygen transfer, against 35% in bottles without it, according to Chanut and colleagues. At 20°C the barrier held steady from three months to 24. At 50°C they recorded "a tremendous oxygen transfer occurring at the glass-cork interface" from three months, and by six months the total diffusion coefficient had climbed towards that of oxygen in open air. Their stoppers were microagglomerated cork, so read it as a measurement of the joint rather than of a one-piece natural cork.

That is the practical reading. A cellar that cycles is not gently ageing your wine early. It is loosening the joint between cork and glass, which is where seepage stains come from. The AWRI's own recommendation is insulated or temperature-controlled storage "which minimise fluctuations in both temperature and humidity", monitored with a logging device rather than trusted. If your storage is a garage or a loft, read why swings do more damage than steady heat before you spend anything on a cabinet.

What humidity does a wine cellar need?

Between 60 and 80%, with less precision than anyone selling humidifiers implies.

The Comité Champagne publishes 60 to 80% as its range. Jancis Robinson notes that 75% is often cited as the ideal, and adds the caveat that matters: little research exists to establish an optimal range.

Both ends have a real failure mode. Too dry and a natural cork loses moisture, which is why the AWRI notes that "wines under natural closures can dry out and leakage can occur if the air is too dry". Too damp and you lose labels and cases to mould, which costs you at resale rather than in the glass.

Experts disagree on how much any of it matters. Matt Kramer, writing in the Wine Spectator, cited a French study claiming relative humidity inside a bottle stays at 100% whatever the closure or the bottle's orientation. If that holds, the cork's wine-facing end is never at risk and only the exposed end is in play. Alexis Lichine took the opposite view and recommended spreading half an inch of gravel on a cellar floor and sprinkling it with water.

One number cuts through the argument. Cork producers specify the stopper's own moisture content, and Relvas Cork's technical sheet for its sparkling closures gives 6% plus or minus 3%, measured to ISO 9727-7, with the corks held at 15 to 25°C and 50 to 70% relative humidity for the 24 hours before use. A cork is not a sponge that needs topping up. It is a component with a moisture spec, and cellar humidity exists to keep it inside that spec rather than to soak it.

The practical consequence is about equipment. Domestic refrigeration dehumidifies as it cools, which is why Tom Stevenson warns against keeping wine in a kitchen fridge for any length of time. That is a real reason to buy a wine cabinet rather than use a spare fridge, and the wine storage humidity guide works through what to do in a dry flat.

Does light reach wine in a cellar?

In a proper cellar, no. In a rack in a living room, yes, and faster than the timescales collectors assume.

The AWRI's technical note sets out the chemistry, crediting Maujean and Seguin's 1983 work: lightstruck flavour, the goûts de lumière, comes from volatile sulfur compounds believed to derive from the sulfur-containing amino acids methionine and cysteine. Riboflavin, present at low levels in musts and wines, "undergoes photo-activation and subsequent reduction when exposed to light at wavelengths of 370 nm and 440 nm", and the reactions that follow build the compounds responsible.

The timings come from Dozon and Noble's 1989 trial, which placed bottles 35 cm from two 40 watt fluorescent lamps. In green glass, the fault reached statistical significance after 31.1 hours for a still white wine and 18 hours for a sparkling one. In clear glass the same wines got there in 3.3 and 3.4 hours. Tasters recorded less citrus character and more "cooked cabbage", "corn", "wet wool/wet dog" and "soy/marmite". Read those hours as a worst case rather than a cellar forecast: 35 cm is closer than any bottle sits to a real light fitting, and the AWRI notes that illuminance falls with the square of the distance.

Sunlight is a different order of magnitude again. The AWRI's comparison: a 36 W fluorescent lamp at 400 lux delivers around 14 mW/m² of UV-A, while full sunlight delivers 60,000 mW/m². The AWRI states that direct sunlight provides 4,286 times the UV-A of a fluorescent tube. A sunlit windowsill is not a slower version of a lit room.

Glass colour is your only passive defence, and it is partial. Rankine's finding, cited in the same note, is that amber glass is most effective at excluding wavelengths below about 450 nm, which is why the risk ranking runs clear, then green, then amber. The AWRI adds a second consequence for white wines: light exposure worsens copper instability, and above roughly 0.5 mg/L of copper "protection from exposure to light will only delay the inevitable". Cases and cartons solve the domestic version of this problem for nothing. The wine storage light guide covers cabinet doors, LED strips and display racks.

Does vibration damage wine?

One study says a little, over 18 months, at intensities a domestic cellar rarely reaches. That is the whole evidence base, and it should be read as one result rather than a rule.

Chung, Son, Park, Kim and Lim, in the Journal of Food Composition and Analysis in 2008, stored a commercial red wine for 18 months at four vibration intensities: 1, 5, 10 and 20 Gal. pH held at 3.4 throughout, according to Chung and colleagues. Total acidity ran between 5.8 and 6.5 g/L, rising slightly up to nine months, and rising most at 20 Gal. Tartaric acid, succinic acid and tannin concentrations fell. The authors' summary: "The changes in organic acids, tannins, and refractive index were more accelerated when the wine was stored at higher levels of vibration than at lower levels of vibration."

Set that against what the AWRI could find on the most-repeated domestic version of the claim. On the argument that a refrigerator compressor's vibration harms stored wine, the AWRI reports in that technical note that "the Institute has been unable to find any literature on this subject".

So the defensible position is the Comité Champagne's: keep bottles away from vibration alongside light and odours, because it costs nothing to do. It is not a reason to move a collection, and it is nowhere near a reason to accept a warmer room in exchange for a quieter one. The wine storage vibration guide goes through washing machines, underfloor plant and living above a railway line.

Should bottles lie down or stand up?

Lie them down, but hold the conviction loosely. The measured differences are small.

Skouroumounis and colleagues gave the most useful data in 2005, tracking a wooded Chardonnay and a Riesling across 60 months. Testing yellow-brown colour at OD420nm, ascorbic acid and sensory analysis, "the bottle orientation (horizontal or upright) was shown to have little effect on the chemical composition and sensory properties of the two wines". Mas and colleagues had earlier seen upright bottles carrying more yellow-brown colour than horizontal ones at 24 months, but the differences were not significant. Chanut's oxygen work agrees from the other direction: over 24 months at 20°C, storage position had no significant influence on oxygen transfer, through the cork or at its interface with the glass.

One difference in the Mas study did hold up, and it is about the closure rather than the angle. Upright bottles sealed with agglomerated cork oxidised more, with raised acetaldehyde in the white wine from the three-month point onwards.

Champagne is the interesting exception, and the current guidance is not what the older books say. The Comité Champagne's own advice is that bottles can be stored upright or on their sides, in a well ventilated place at 10 to 15°C and 60 to 80% humidity, protected from light, vibrations and odours. The guide to storing bottles on their side works through racking, cases and the slight-angle argument.

The reason horizontal remains the default is not the chemistry. It is that a bottle on its side keeps the cork wet, and a wet cork is the one thing you cannot inspect from outside the case.

Know when the cellar's work is finished

Good storage buys time. The point of the time is to open the bottle at the right moment, and that is the number most collections get wrong, because a cellar log records what you own rather than what is ready.

Every wine here carries a drink window built from critic windows and CellarTracker drinking velocity, so you can see when a vintage moves from holding to drinking and how fast the crowd is getting through it. A first growth like Château Latour and a wine like Romanée-Conti do not run on the same clock, and neither runs on the clock a merchant's release note gave them.

Join the newsletter and we will tell you when a wine you own enters or leaves its window, before the market works it out. Start with wines drinking now for what is ready this season, the market index if you want the price side of the same decision, or the Bordeaux producer atlas if you are still deciding what deserves the cellar space.