Does the type of water you add to whisky actually change the taste? What the evidence says

Key takeaways

• In the one controlled trial to test it directly, water type made no significant difference to the taste and mouthfeel of aged spirits. Spring water, distilled water and limestone water all proofed a whisky to 40% ABV without a trained panel reliably telling them apart (Journal of Sensory Studies).
• That null result is not a licence to pour anything. Minerals and contaminants live on completely different scales. Calcium becomes tasteable around 100–300 mg/L; the chlorophenols formed when chlorine meets phenols are tasteable at 0.1 µg/L — roughly a million times lower (WHO).
• What demonstrably changes a dram is how much water you add, not which brand. Dilution reorganises the liquid at a molecular level, driving aroma compounds such as guaiacol to the surface (Scientific Reports) and shifting seven of ten sensory descriptors in a 20-person trained panel (Foods).
• The practical rule survives the science intact: use still, low-mineral, unchlorinated water at room temperature. Not because minerals will be tasted, but because the things that will be tasted travel in ordinary tap and mineral water.

Yes — but far less than the internet claims, and almost never for the reason given. When researchers at the University of Illinois diluted aged and unaged spirits down to 40% ABV with three different waters and put them in front of a trained sensory panel, water type had no significant effect on the aged spirits at all. The differences drinkers genuinely notice come not from a water’s minerals but from its contaminants — chlorine and its by-products, iron, and the calcium that can throw a permanent haze — which are detectable at concentrations thousands of times below where minerals start to register.

What does the research actually show?

Until recently, almost nothing had been tested. Water is the second ingredient in every bottle of whisky, and yet, as Zhuzhu Wang and Keith Cadwallader put it in the Journal of Sensory Studies, “its impact on the sensory perception of distilled spirits after proofing is scarcely reported.”

Their 2024 study closed part of that gap. They took aged and unaged spirits above 55% ABV and proofed them down to 40% ABV using three waters — spring, distilled, and limestone — and two methods, fast and slow. A trained panel then assessed the results for taste and mouthfeel.

The headline finding: “water type and proofing method did not have a significant effect on aged spirits.” For unaged spirits, the panel did detect something — but it was the method, not the water. Slow-proofed unaged spirits “were characterized as sweeter at the beginning, with less burn throughout the entire evaluation period” (Journal of Sensory Studies).

Two caveats keep this honest. The study compared three potable waters — no chlorinated tap water was in the line-up — and it measured taste and mouthfeel, not aroma. It is strong evidence that ordinary mineral variation is inaudible in an aged spirit. It is not evidence that any liquid labelled water will do.

Then why do so many tasters swear they can tell?

Because informal tastings are not blind, and because some of the waters people compare really do carry flavour of their own.

The most-cited demonstration is a 2016 tasting for the Tales of the Cocktail Foundation, in which three bourbons were sampled against four waters: Old Limestone, a high-mineral well water, tap water, and distilled water. Distilled water had “limited effect on the same bourbon (dilution without sensory changes).” The mineral-rich well water “brought out floral aromas but was notably more brittle in the mouth.” And the tap water “added notes of swimming pool and band-aids” (Tales of the Cocktail Foundation).

That last note is the one to hold on to. It is not a mineral effect — it is a chlorine effect, and it is the most reproducible water fault in whisky tasting. Read the experiment as a whole and it broadly agrees with the controlled study: the neutral waters behaved neutrally, and the outlier was the treated one.

Something real sits behind the mineral observations too. Rachel Barrie, master blender at Bowmore, notes that “soft water (like Speyside and Kentucky limestone water) is softening on the palate and can bring out fruit and honey notes.” Whether that survives blind conditions is exactly what the trained panel tested — and it did not. Our companion piece on what master distillers actually say about water covers how wide that disagreement runs inside the trade.

How much water is actually in the glass?

This is the arithmetic that makes the null result intuitive.

Pour 30 ml of a 60% ABV whisky and add 10 ml of water, and water is 25% of what is in the glass — a heavy dilution by most drinkers’ standards. Now take a genuinely mineral-rich water. Evian publishes a dry residue of 345 ppm (mg/L), with calcium at 80 mg/L, magnesium at 26 mg/L and bicarbonates at 360 mg/L (evian).

At a quarter of the glass, that water contributes roughly 20 mg/L of calcium to the final liquid. The WHO’s guidance on the acceptability of drinking water puts the position plainly: “The taste threshold for the calcium ion is in the range of 100–300 mg/l” (WHO).

So the calcium arriving from a well-known mineral water lands roughly five to fifteen times below the concentration at which calcium becomes tasteable in water at all — before accounting for the fact that it is now competing with a spirit carrying hundreds of congeners. The other minerals fare no better: sodium’s average taste threshold is about 200 mg/L and sulfate’s runs from 250 mg/L upward, while this water delivers 6.5 mg/L and 15 mg/L respectively. The trained panel’s inability to distinguish spring from distilled water is not a surprise; it is what the numbers predict.

Which thresholds actually matter, then?

The ones that sit several orders of magnitude lower. This table is the real answer to the question in the title.

Between the top of that table and the bottom lies a factor of about a million. Minerals in drinking water are, for practical purposes, below the horizon. Chlorine chemistry is not.

The mechanism matters here. Chlorophenols form through the “reaction of chlorine with phenols” during water treatment, and whisky — peated whisky above all — is a phenol-rich liquid. Adding chlorinated water to it is the one intervention in this whole debate where a genuinely potent off-flavour class is even theoretically in play, and the “swimming pool and band-aids” note from that bourbon tasting is precisely what chlorophenol taint smells like.

Run the dilution arithmetic again and the asymmetry is stark. A tap water carrying 1.0 mg/L of free chlorine, at a quarter of the glass, still delivers 0.25 mg/L — right at the edge of WHO’s stated detection floor. The calcium in the same glass sits an order of magnitude below its own.

One honest limitation: these thresholds are established for water, not for 40% ABV spirit. Ethanol changes both solubility and perception. They are the correct order-of-magnitude guide, not a measurement of your dram.

Do minerals ever change the taste of a spirit?

Yes — at concentrations no drinking water comes close to.

Research on Japanese pot-distilled spirits found that specific salts can soften what the literature calls the “alcohol-stimulative taste” — the burning, irritative edge of ethanol. Akira Nose reported in MOJ Food Processing & Technology that MgCl₂, MgSO₄ and NaHCO₃ each reduced that stimulative character, while CaCl₂, NaCl, Na₂SO₄ and KCl did not, with ¹H NMR indicating the effective salts “could strengthen the hydrogen-bonding structure of water-ethanol” (MedCrave).

It is a genuinely interesting result, and it is routinely misquoted as proof that mineral water improves whisky. Look at the doses. The effective salts were tested around 0.5 M — for magnesium chloride, on the order of 12,000 mg/L of magnesium, against the 26 mg/L Evian delivers. The experiment used magnesium at roughly four hundred times the concentration of a mineral water, in a laboratory ethanol solution, to move a single attribute. The mechanism is real; the dose explains why nobody meets it in a glass.

What definitely changes the taste is how much water, not which

Set the brand question aside and the science on dilution itself is unambiguous.

Molecular dynamics simulations by Björn Karlsson and Ran Friedman showed that guaiacol — a key smoky, vanilla-like aroma compound — “is preferentially associated with ethanol and primarily found at the liquid-air interface” in mixtures up to about 45 vol-% ethanol, while at 59 vol-% or above it is “increasingly surrounded by ethanol molecules and is driven to the bulk” (Scientific Reports). Dilution literally moves the aroma to where you can smell it.

Sensory work confirms the effect at the glass. A 2023 study in Foods examined 25 whiskies chemically and put six in front of a 20-person trained panel at four dilution levels. Dilution reduced hydrophilic compounds in the headspace while increasing hydrophobic ones; seven of ten sensory descriptors differed significantly, with undiluted samples pulling toward “Solvent/Chemical” attributes and diluted samples toward “Pome Fruit” and “Cedar” (Foods).

That is the real lever. Our water-to-whisky ratio guide by style and our drop-by-drop dilution guide cover how to use it.

Is there a case against mineral water that has nothing to do with taste?

There is, and it is the one distilleries actually act on. Calcium in dilution water reacts with oxalic acid from the oak to form calcium oxalate — an irreversible haze that never redissolves. As little as 2 ppm of calcium in the finished spirit can trigger it, which is why commercial producers proof exclusively with demineralised water rather than trusting a spring (Artisan Spirit Magazine). We unpack that in why whisky goes cloudy when you add water and set out the target specification in demineralised water specifications for spirit dilution.

At the distillery’s scale, then, water type is not a matter of taste preference at all — it is a physical-stability requirement with a 2 ppm tolerance. The trade’s near-universal use of demineralised water was never a claim that mineral water tastes wrong. It is a claim that mineral water is a risk.

So what should you actually pour?

The recommendation is unchanged by the research; only the reasoning is sharper.

• Still, low-mineral, unchlorinated water. Chlorine is the one thing in ordinary tap water whose detection threshold survives dilution into a glass. If you use tap water, that is the variable to eliminate.
• Don’t chase the “right” mineral profile. A trained panel could not distinguish spring from distilled water in aged spirits. Buying a specific mineral water for its calcium is optimising a variable that sits below its own taste threshold.
• Avoid high-bicarbonate waters if you want zero risk. Writing in Scotch Whisky, Felipe Schrieberg argues water for whisky “should be as neutral as possible in taste,” keeping bicarbonates below 150 mg/L because they “can be astringent” — and singles out Evian’s 360 mg/L as too high (Scotch Whisky). Treat that as a well-reasoned trade opinion rather than a measured result, but nothing is lost by heeding it.
• Room temperature, never chilled. Cold suppresses volatile aromas and encourages haze — it works against you twice.
• Add it slowly. The one variable the 2024 study did find significant was proofing method, not water type.

For the fuller comparison of what sits within reach of most drinkers, see tap vs filtered vs mineral vs demineralised water, and for reading a spec sheet, pH and TDS explained.

Glossary

• Taste threshold — the lowest concentration at which a substance becomes detectable by taste. Published thresholds are usually established in water, not in spirit.
• TDS (total dissolved solids) — the combined mass of everything dissolved in water, in mg/L. Often reported as “dry residue at 180°C” on mineral-water labels.
• Alcohol-stimulative taste — the burning, irritative sensation of ethanol, treated as a measurable sensory attribute in spirits research.
• Chlorophenols — compounds formed when chlorine reacts with phenols; responsible for medicinal, “band-aid” and swimming-pool off-flavours at sub-microgram-per-litre concentrations.
• Guaiacol — a smoky, vanilla-like aroma compound in whisky, and the molecule most studied in relation to dilution because it migrates to the liquid surface as ethanol falls.
• Proofing (fast vs slow) — adding dilution water in one go versus incrementally. The one process variable shown to significantly change the sensory profile of unaged spirits.
• Calcium oxalate — the insoluble salt formed when calcium in dilution water meets oxalic acid from the cask. The industry’s practical reason for using demineralised water.
• Demineralised water — water stripped of dissolved minerals by reverse osmosis and/or ion exchange; the industry standard for proofing spirits.

Frequently asked questions

Does the type of water change the taste of whisky? In controlled testing, barely. A trained panel found no significant taste or mouthfeel difference between aged spirits proofed to 40% ABV with spring, distilled or limestone water. The waters that reliably change a dram are chlorinated or otherwise contaminated ones, not differently mineralised ones.

Why does tap water make my whisky taste medicinal? Almost certainly chlorine. People can detect free chlorine in water at concentrations as low as 0.3 mg/L, and the chlorophenols formed when chlorine meets phenolic compounds are detectable at 0.1 µg/L. Whisky is phenol-rich, peated whisky especially so.

Is mineral water bad for whisky? Not in taste terms at ordinary concentrations — the minerals sit well below their own taste thresholds once diluted into a glass. The real objection is physical: calcium can form an irreversible haze with oxalic acid from the cask, and distilleries treat 2 ppm as the limit.

Should I use distilled water instead? It works, and in the Illinois study it was indistinguishable from spring water in aged spirits. Demineralised or distilled water simply guarantees that the water contributes nothing at all — no chlorine, no minerals, no haze risk.

Does the amount of water matter more than the type? Yes, decisively. Dilution changes which aroma compounds reach the surface of the liquid and shifted seven of ten descriptors in a trained-panel study. Water type moved nothing measurable in the same aged spirits.

Do distilleries use mineral water to proof whisky? No. Commercial producers reduce spirit with demineralised water, made on site or supplied to specification, precisely to avoid contributing minerals. The famous local spring belongs to the mash, not the final cut.

Blenders International · Water Science series