A corked wine is one contaminated by 2,4,6-trichloroanisole (TCA), a compound that forms when airborne chlorine meets naturally occurring mold in cork bark. At concentrations as low as 1.5 to 3 nanograms per liter, TCA numbs the nose and delivers the signature smell of wet cardboard, dank basement, or musty newspaper. The wine itself is not harmful to drink. "Corked" describes a sensory defect, not a safety problem.
Cork taint is the most studied fault in wine, and also the most misdiagnosed. The compound responsible operates in parts per trillion, which is a scale most drinkers cannot visualize. That tiny dose is enough to suppress the aromas the wine was made to carry. Here is what is happening at the molecular level, how common the fault really is, why experienced tasters still get it wrong, and what you can reasonably do about it.
The molecular route from chlorine to TCA
The pathway starts in the forest, not the cellar. Cork oak bark hosts a diverse community of filamentous fungi. Those fungi are not harmful in themselves, and the bark is harvested every nine years without killing the tree. But when the bark is bleached or washed with chlorinated water, or when ambient chlorine compounds from the environment settle on it, the fungi methylate those chlorophenols into chloroanisoles. TCA is the dominant one, though related compounds like 2,4,6-tribromoanisole (TBA) and pentachloroanisole contribute to the broader category of "haloanisole" taint.
The mechanism matters because it explains why the fault is so persistent. TCA is somatically active at vanishingly low concentrations. Human olfactory receptors detect it around 1.5 to 3 nanograms per liter in water, and the threshold in wine is similar or slightly higher depending on the matrix. One gram of pure TCA could theoretically contaminate an enormous volume of wine at threshold level. The molecule is also lipophilic and stable, so it survives bottling, storage, and—this is the counterintuitive part—does not reliably disappear with time in the bottle. Some tainted wines fade; many do not.
TCA does not simply add a smell. It suppresses other aromas by interfering with olfactory signaling, which is why a corked wine often reads as "muted" or "flat" before it reads as "musty." The damp-cardboard character emerges at higher concentrations, roughly 5 to 10 ng/L and up. This dual action is why the fault is so easy to miss in lightly tainted bottles.
How often it happens, and what the numbers mean
Historical estimates put cork-taint incidence in bottled wine at roughly 2 to 7 percent of natural-cork closures, with some survey ranges wider. Those numbers come largely from producer and industry audits rather than a single gold-standard prevalence study, so treat them as a working range rather than a precise rate.
The economic picture is instructive. A producer bottling 100,000 cases with a 3 percent taint rate faces roughly 3,000 affected cases—far more than most wineries can afford to absorb silently. That pressure drove an industry-wide shift toward alternative closures over the past two decades. Screwcaps and synthetic corks reduced the problem but introduced their own issues, notably reduction (sulfur-based) faults and, in some synthetic closures, a different taint profile.
Analytical testing has improved. Producers now screen cork lots with gas chromatography-mass spectrometry and can reject batches above a set threshold. No screening program is perfect, because taint can be localized within a single stopper and sampling all corks is impossible. A wine can therefore be tainted even when its closure tested clean.
Timeline: development of taint and how it presents
| Stage | What happens | Practical signal |
|---|---|---|
| Cork processing | Chlorinated wash or ambient chlorine contacts fungal growth | Lot-level risk set before bottling |
| Bottling to first month | TCA partitions between cork and wine | Aroma may already be muted |
| 1–6 months | TCA equilibrates in the headspace and liquid | Musty note becomes more detectable |
| 6–24 months | Compound stable; no meaningful degradation | Light taint persists, not "blows off" |
| 2+ years | Slow decline in some bottles, none in others | Unreliable expiry; do not cellar and hope |
The table captures a professional caveat: some tainted wines do fade with long aging, but there is no reliable way to predict which. Advancing the timeline is not a fix.
Why tasters disagree, and the counterintuitive part
TCA detection is genetically variable. Some people carry olfactory receptor variants that make them far more sensitive to haloanisoles; others can barely detect them. Age, smoking status, recent meals, and even the wine's own structure (high tannin and acidity can mask the musty note) shift the reading. Two careful tasters can honestly disagree about the same bottle.
The counterintuitive detail is that a tainted wine is often judged "off" rather than "faulty." Beginning blind tasters describe a corked red as thin, leathery, or dried out—without naming the actual fault—because the suppression of fruit and florals dominates. The musty smell is a later, higher-dose symptom. This is why professional panels train with spiked reference samples rather than relying on bottle-by-bottle intuition.
There is also a storage variable. Many "corked" complaints are actually oxidation or storage damage, especially in wines kept near heat or under poor humidity. A genuinely musty cabinet or basement can transfer haloanisoles to wine through the closure—so the cork can pick up taint from the air after bottling, not just the winery.
Comparison: corked versus other common faults
| Fault | Signature | Cause | Harm? |
|---|---|---|---|
| Cork taint (TCA) | Wet cardboard, damp basement, muted fruit | Haloanisoles from cork mold | No |
| Oxidation | Sherry-like, nutty, brown-edged | Oxygen ingress or heat | No |
| Reduction | Struck match, rotten egg, burnt rubber | Low oxygen, sulfur compounds | No |
| Brettanomyces | Barnyard, band-aid, horse | Wild yeast | No |
| Volatile acidity | Vinegar, nail polish | Bacterial acetic acid | No at low levels |
| Heat damage | Stewed fruit, leaks | Elevated storage temperature | No |
| TBA taint | Musty cardboard, no chlorine source | Wood pallets or packaging | No |
The table shows why the "musty" label alone is insufficient. Cork taint is chemically distinct from oxidation and reduction, and the recommended responses differ.
Practical actions: what to do with a suspect bottle
Smell the cork and the wine separately before deciding. The cork itself can carry a faint musty note that may not have reached the wine. Pour a small sample, swirl, and wait 30 seconds. TCA's signal intensifies with air contact rather than fading. Compare against a known-good reference wine if possible. Blind comparison is the most reliable home test. If the wine is served at a restaurant, say clearly that it seems corked and request a replacement. Most restaurants have a policy for this, and the fault is not your error. Contact the producer or retail channel with the bottle, closure, and lot code for a refund or replacement. Major producers track these reports. Do not cook with a tainted wine. The compoound carries through and can spoil a sauce.
Red-flag warnings for distinguishing true taint from disappointment: The wine is muted and lacks fruit but shows no musty note—could be reduction or a flawed bottle, not necessarily TCA. The odor is vinegar or nail polish—this is volatile acidity, not cork taint. The wine is brown and nutty—oxidation or heat, not TCA. The wine was stored near a musty basement or cardboard pallets—environmental pickup possible. You smell "wet dog" but the bottle is a screwcap—suspect TBA rather than cork taint.
The reconciliation point: commercial thresholds for rejection commonly sit near 1 to 2 ng/L in water-based reference tests, while sensory threshold data cluster around 1.5 ng/L, and higher adverse detection near 5 to 10 ng/L. These are not competing numbers; they describe detection, recognition, and rejection at different points on one curve. No single number defines "corked" in a legal sense. The judgment is sensory.
Limitations and how far the science goes
Prevalence data are estimates, not census counts. Detection thresholds vary by individual and by wine matrix, which is why laboratory screening and human sensory panels sometimes disagree. Analytical instruments solve different problems than panels do: a GC-MS screen can quantify TCA but cannot tell you whether a specific bottle crosses the buyer's perception threshold.
Taint also has more than one source. Haloanisoles can come from cork, wood pallets, air conditioning ducts, and chlorine in the processing environment. Blaming every musty wine on the cork is inaccurate, and modern production has shifted much of the risk away from the closure itself in some supply chains.
There is no consumer-safe way to "fix" corked wine. Anecdotes about plastic wrap or decanting are inconsistent and unsupported by controlled evidence. The honest answer is that the defect is not reversible at home, and the remedy is replacement, not repair.