Wine Oxidation: How to Spot It and How to Stop It

Wine oxidation rarely announces itself. A bottle can be filled to the neck, sealed and stored at the right temperature, and still pour flat and gold because the sulphur dioxide protecting it ran out two years ago. Oxidation happens when oxygen reaches wine faster than its own defences absorb it. Bueno and colleagues, in Frontiers in Chemistry in 2018, name the culprits: "once oxygen is dissolved in wine, its activation is carried out by metal ions, particularly iron and copper in lower oxidation states." That chain oxidises ethanol into acetaldehyde. The Australian Wine Research Institute reports the sensory threshold for acetaldehyde at 100 to 125 mg/L, and above 125 mg/L names "over-ripe bruised apples", "stuck ferment" character, or "sherry" and "nut-like" characters. But you smell the damage long before acetaldehyde arrives. The same study found Strecker aldehydes "will accumulate before acetaldehyde", with methional and phenylacetaldehyde reaching odour activity values "above 20, when acetaldehyde is barely detectable". Three things decide it: temperature, the closure, and time.

What does an oxidised wine smell and look like?

It smells tired before it smells like sherry. The AWRI describes a ladder that runs "from a dulling of the aroma, to 'cardboard', 'straw' and 'hay-like' aromas, to 'sherry-like' and 'madeirised'", and in extreme cases a "'wet wool', 'wet dog', or 'varnish-like' aroma".

Note where that ladder ends. Sherry-like and madeirised name two wines where the oxidation was the whole intention. In a Puligny it is a fault. In an oloroso or a Madeira it is the point, and the chemistry does not care which bottle it is in.

The colour moves first and it moves in one direction. The Oxford Companion to Wine, quoted by The Real Review, describes a prematurely oxidised white as showing "some advance in the colour, more in the dull yellow range", with aromas of "bruised apples or furniture polish". Whites deepen towards gold and brown, reds lose purple at the rim and go brick then tawny. Our colour and age guide sets out what each shade is normal for.

There is one field test worth knowing. The Oxford Companion adds that those aromas "can strengthen considerably with aeration". That is the opposite of a reductive wine, which is why the glass is a diagnostic instrument and not just a serving vessel. Leave both in the glass for twenty minutes: the reduced wine opens, the oxidised one gets worse.

Why does a sealed bottle oxidise at all?

Because the seal is not a seal, and the wine's defence runs down on a schedule.

Chanut and colleagues, publishing in PNAS Nexus in 2023, tracked oxygen into bottles over 24 months and found the weak point is not the cork body but the joint between cork and glass. Studied dry, that interface accounted for 35% of the total oxygen transfer. With model wine present, it accounted for nearly 75%. Their explanation is mechanical: the stopper hydrates and plasticises, so "the force applied to the glass surface walls of the bottleneck by the stopper decreases as the material is hydrating". Oxygen diffusion through the bottleneck and stopper was already higher after three months.

Meanwhile the sulphur dioxide that scavenges the reaction is being spent. Cantu and colleagues at UC Davis bottled a Napa Sauvignon Blanc under three closures and followed it for 30 months, publishing in Molecules in 2022. Free SO2 measured 36 mg/L at bottling. One month later it was 23 mg/L. At 30 months it sat between 5 and 12 mg/L depending on the closure.

So the honest model of a maturing bottle is a race. Oxygen arrives at a rate set by the closure and the temperature. Sulphur dioxide depletes at a rate set by how much oxygen arrives. The wine is protected until the antioxidant runs out, and then it is not, and the change is quick. That is why premature oxidation shows up as a cliff rather than a slope, and why one bottle in a case can fall off it while the other eleven do not.

How do you tell oxidation from heat damage, cork taint and reduction?

Four faults, four signatures, and only one of them browns the wine.

• Oxidation. Colour advances. Fruit dulls, then bruised apple, then sherry. Worsens in the glass with air. AWRI descriptors above.
• Heat damage. Often arrives with oxidation but leaves physical evidence first: seepage under the capsule, a pushed cork, ullage that runs ahead of the bottle's age. The AWRI warns that temperatures "in excess of 40°C will induce visual and sensory changes to a wine in only a matter of days". Our ullage levels guide covers how to read the fill line against the vintage.
• Cork taint. Musty and mouldy, and it subtracts rather than adds. The AWRI describes 2,4,6-trichloroanisole as having "a distinct musty, mouldy aroma", puts its aroma threshold in a Pinot Noir at 1.4 ng/L, and notes the threshold "might be lower in dry, white wines". A corked wine is pale and mute, not gold and nutty.
• Reduction. The mirror image. The AWRI lists the detection threshold for hydrogen sulfide at "about 1 – 2 µg/L", ethyl mercaptan at 1.1 µg/L for "'onion-like' and 'rubber-like'" aromas, and methyl mercaptan at "0.02 – 2.0 µg/L" for "'rotten eggs' and 'cabbage'". Reductive character blows off with air. Oxidation does not come back.

Get this wrong at the table and you will decant a corked bottle hoping it improves, or dismiss a merely reduced wine that needed ten minutes.

What was premox, and is white Burgundy fixed?

Premature oxidation is the case that taught the fine wine market that a sound-looking bottle can already be dead. World of Fine Wine credits Steve Tanzer with raising it publicly in 2002, in International Wine Cellar. "The afflicted vintages are predominantly from the late 1990s, and in particular those of 96, 97 and 98, [ 3 ] until."

Clive Coates MW dates the onset to "pre-1995, the vintage when all the problems seemed to begin", names 1996, 1997 and 1998 as worst, and records that "the incidence of 'prem-ox' bottles was and continues to be distinctly higher in the USA than in France and Great Britain". His own examples are the ones that stick. At an Atlanta dinner his table opened a full case of Puligny-Montrachet Les Combettes from Domaine Etienne Sauzet to get four decent bottles. Of his dozen 1999 Puligny-Montrachet Les Demoiselles from Domaine Colin Déléger he tried three: "All excessively oxidised before their time." On who was hit, he names Lafon, Bonneau du Martray, Roulot and Ramonet, and writes that "no one seemed to be spared".

The cause was argued for a decade. Coates held that "poorly-performing corks are the main culprits behind prematurely aged white Burgundy". Pierre Rovani of The Wine Advocate countered that "corks are not the issue". Allen Meadows split the difference: "the most likely source of the problem is cork-related, though it appears this has been exacerbated by generally lower levels of SO2." Tanzer's list ran wider, taking in corks, global warming producing overripe fruit, excessive lees stirring and insufficient sulphur dioxide. Denis Dubourdieu and Valérie Lavigne in Bordeaux proposed that vine stress left grapes with "insufficient quantities of glutathione, a compound that functions as an essential antioxidant".

The sulphur numbers explain a lot. World of Fine Wine reports that Burgundian levels were reduced to perhaps 15-20 parts per million (ppm) before bottling from the mid-1990s, against 30 to 40 ppm in more recent practice, with around 15 ppm eaten up by the bottling process itself. Dominique Lafon had problem bottles analysed and found total sulphur down from 100ppm at bottling to 50ppm, which is impossible unless some agent has eaten the sulphur.

Two things followed. Producers raised sulphur, scrutinised cork quality and cut back on bâtonnage. And several of the best-known names left natural cork entirely. Jamie Goode records that Domaine Leflaive announced on 31 October 2016 that it would bottle its whole range under DIAM beginning with the 2014 vintage, having suffered badly from the 2006 vintage onwards, with Dominique Lafon on DIAM from 2013 and Benjamin Leroux on screwcap for all his whites from 2014. Diam's own commercial director told The Real Review that DIAM now seals 60% of grand cru white Burgundy and 45% of all Burgundy, though that count dates from 2020 rather than this year.

So yes, the problem has been addressed for wine bottled since. It has not been addressed for wine already in your cellar. A Burgundy 2015 white carries a different closure risk from a 1999, and nothing done in the winery in 2016 reaches back into a case bought in 2001. The Burgundy producer atlas shows which estates sit where, but the closure decision, and the vintage it started from, is the number that matters for an old case.

Which bottles in your cellar are most at risk?

Whites, old wine, and anything whose fill line has moved.

The AWRI is direct about grape variety: white wines made from the "floral" varieties such as Riesling are "very prone to oxidation", whereas red wines "can withstand significant oxidation during handling due to the higher content of phenolic compounds, which are natural antioxidants". Age compounds it. In the Frontiers study, samples from young red wines held acetaldehyde "under 19 mg L −1, regardless of the O 2", while samples from aged wines rose "above 25 mg L−1" in the most oxidised group. The old bottle has less buffer left, which is exactly when people start opening it.

Do not assume sweetness or reputation is a shield. Dubourdieu and Lavigne, quoted in World of Fine Wine, concluded that premox affects all white wines, still and sparkling, dry and sweet, and all grape varieties and origins. A fifty-year drink window on something like Château d'Yquem is a statement about the wine's structure, not a guarantee about the closure in front of it.

Does the closure decide it?

The closure decides the spread more than the average, and the spread is what costs you.

Cantu and colleagues measured browning at 30 months across three closures on the same wine: 0.84 for synthetic cork, 0.81 for natural cork, and 0.89 for screw cap. Those means are close together. The variability was not. Natural cork came in at a coefficient of variation of 44.8%, against 21.9% for synthetic and 20.6% for screw cap, and the screw caps were specified by their maker at an oxygen transmission rate of 0.5 mg/year.

Then the finding that should keep everyone honest. Their trained descriptive panel of eleven tasters "could not detect any difference in their aroma and flavor profile between closure types regardless of browning level", and that held down to the most browned natural-cork group, which finished the study at 5 mg/L of free SO2. Measurable oxidation and perceptible oxidation are not the same thing, and an instrument reading is not a verdict on a bottle.

The practical reading is about variance, not means. A case under natural cork is twelve individually different seals, which is precisely the pattern Coates found in his cellar and that Michael Brajkovich MW pointed to when he told The Real Review that if vineyard factors were the cause, "then surely all of the wines in that dozen would have been similarly affected". DIAM and screwcap exist to remove that lottery. Neither makes a wine better. Both make it more predictable, which for anyone holding a case as an asset is the more valuable property.

What storage actually stops oxidation?

Cool and steady, and the numbers are published rather than folklore.

The AWRI recommends bottled wine be stored "with the cork in contact with the wine in a cool (15–20°C), dry location", and warns that "any storage place where the temperature exceeds 25°C for long periods and 40°C for short periods can affect wine quality". It adds that "thermal cycling where the temperature varies significantly between day and night should be avoided", and that wines under natural closures "can dry out and leakage can occur if the air is too dry".

Temperature matters because it drives reaction rate. Venturi and colleagues, writing for IntechOpen in 2019, put it through the Arrhenius relationship: "as the temperature rises, there is an increase in the rate of occurring reactions", and "high temperature is a particularly unfavorable condition during storage as the rate of quinone formation enhances with the increase in temperature". Their saturation figures for oxygen run "from 10.5 to 5.6 mg/L" across 5°C to 35°C, which corrects the idea that warm storage is safe because warm wine holds less oxygen. It holds less and consumes it faster.

Heat also attacks the seal itself. Chanut's group saw that at 50°C "a tremendous oxygen transfer occurring at the glass-cork interface was noticeable, even from 3 months".

One finding worth holding lightly. Across 3 to 24 months at 20°C, the same study found "the storage position of the bottle had no significant influence on the oxygen transfer, neither through the cork (as previously reported) nor through the glass-cork interface". That is two years, not twenty, and the AWRI still asks for the cork in contact with the wine. Lay bottles down. Just stop treating position as the variable that decides oxidation, when temperature stability is. The full set of limits sits in our cellar conditions guide.

What does an advanced bottle cost you at auction?

Oxidation is priced twice: once in the estimate when a house catalogues an old white, and once in your glass when a lot arrives sound-looking and dead.

The open question is whether the market has already repriced the risk or is still pricing white Burgundy by vintage reputation alone. The market index shows the same wines undifferentiated by condition, which is the comparison that makes any adjustment visible, and our methodology sets out how a lot description gets mapped to a condition group.

Open the bottle before the window closes, not after

The only reliable defence against oxidation is drinking the wine while it is still worth drinking. Every wine on this site carries a drink window built from critic windows and CellarTracker drinking velocity, and you can put an alert on the wines you own.

Set it and you get told when a bottle enters the last third of its window, which is the point at which waiting another year stops being patience and starts being a bet. That is worth more than any storage advice, because a perfect cellar only slows the clock. It does not stop it.

Start with drink-now if the question is what to open this month.