What the strength of a cocktail actually depends on
A cocktail's strength is decided by two numbers: how much pure ethanol went into it, and how much liquid that ethanol ends up spread through. The first is fixed by the spec — add up each ingredient's volume multiplied by its own ABV and you have it. The second is not fixed at all, because a large part of the finished drink is water that was never poured. It arrived from the ice.
That is why a Martini and a Negroni built from the same 3 fl oz of liquid taste like completely different propositions. The Martini carries far more ethanol into the mixing glass; both pick up a comparable amount of melt water; the strengths diverge. And it is why the same spec served up, over a single large cube, or over crushed ice reaches the drinker at three different strengths, even though the bartender measured identically each time.
Dilution is the part most people leave out, and leaving it out overstates the strength substantially. A shaken drink typically finishes at around a quarter more volume than what was poured, which pulls the strength down by a fifth — because dividing by 1.25 is multiplying by 0.8. That is not a rounding adjustment, it is the difference between a drink you can serve and one you cannot.
The calculator also converts the ethanol figure into standard drinks. In the United States the National Institute on Alcohol Abuse and Alcoholism defines one standard drink as 14 grams of pure ethanol, which at ethanol's density of 0.78934 g/mL is 17.74 mL. This number is independent of dilution: watering a drink down changes how it drinks, not how much alcohol is in it.
The formula, and why dilution sits in the denominator
Each ingredient contributes ethanol equal to its volume times its ABV. Two fluid ounces of 40% tequila is 59.15 mL × 0.40 = 23.66 mL of ethanol. Sum that over every ingredient and you have the numerator, and it never changes again — nothing you do after pouring adds or removes alcohol.
The denominator is everything in the glass: the poured volume plus the melt water. Dilution is expressed here as a percentage of the undiluted volume, which is the convention bartenders use when they batch, because it is the number you multiply your recipe by. A 25% dilution on 125.7 mL of ingredients is 31.4 mL of water and a served volume of 157.1 mL.
Notice what that does to the strength. Multiplying the denominator by (1 + d) divides the ABV by the same factor, so the relationship between dilution and strength is a reciprocal, not a straight line. Going from 0% to 25% dilution costs you a fifth of the strength; going from 25% to 50% costs you a further sixth of what is left (1.25 ÷ 1.5 = 0.833), not another fifth of the original. This is why the strength column in the reference table below flattens out as you read down.
The one thing the formula assumes is that volumes add. Ethanol and water genuinely contract when mixed, which is why a distiller proofing a spirit down cannot rely on arithmetic alone — see the proof and dilution calculator for that case. At cocktail strengths the contraction is well under the error of a jigger poured by hand, so ignoring it costs nothing.
Worked example: a 2:1:1 margarita, shaken
Spec: 2 fl oz blanco tequila at 40% ABV, 1 fl oz orange liqueur at 40% ABV, 1 fl oz lime juice, and 0.25 fl oz agave syrup, shaken hard with ice.
- Convert to millilitres. One US fluid ounce is 29.5735 mL, so the pours are 59.15, 29.57, 29.57 and 7.39 mL. Total poured: 125.69 mL.
- Add up the ethanol. Tequila: 59.15 × 0.40 = 23.66 mL. Liqueur: 29.57 × 0.40 = 11.83 mL. Lime and agave contribute nothing. Total: 35.49 mL of pure ethanol.
- Undiluted strength. 100 × 35.49 ÷ 125.69 = 28.24% ABV — what is in the tin before the ice does anything.
- Melt water. Shaking at 25%: 125.69 × 0.25 = 31.42 mL.
- Volume in the glass. 125.69 + 31.42 = 157.11 mL, a hair over 5 fl oz, which is why a margarita needs a coupe and not a cocktail glass.
- Finished strength. 100 × 35.49 ÷ 157.11 = 22.59% ABV. Check it against step 3: 28.24 ÷ 1.25 = 22.59, as the reciprocal rule says it must be.
- Standard drinks. 35.49 mL × 0.78934 g/mL = 28.01 g of ethanol, and 28.01 ÷ 14 = 2.00 standard drinks. One margarita is two drinks.
- Batching twelve. 157.11 × 12 = 1.885 L of finished batch, of which 31.42 × 12 = 0.377 L is the water you add yourself.
How to read the result
Most well-made cocktails land somewhere between 15% and 25% ABV as served, which is roughly wine strength — a coincidence that has nothing to do with taste preference and everything to do with the fact that a drink much stronger than that stops being a drink and starts being a measured pour. Spirit-forward drinks such as a Martini or an Old Fashioned run higher, into the mid thirties, because there is almost nothing non-alcoholic in the build. Sours land lower because juice and syrup dilute before the ice does. Highballs land lower still.
The strength number tells you how the drink will taste and feel. The standard-drinks number tells you what it does. Those two come apart in a way that catches people out: a Negroni at 23% ABV contains fewer standard drinks than a Martini at 35%, but a long gin and tonic at 12% can contain the same alcohol as either, because the ethanol is spread through six ounces instead of three. Judge alcohol by the ethanol figure, not by how strong the glass tastes.
If the calculator warns that a serving exceeds two standard drinks, that is a spec decision rather than an error — a 2 fl oz base pour at 40% is already 1.33 standard drinks before anything else is added. The usual levers are a smaller base pour, a lower-proof base, or splitting the base between a spirit and a fortified wine.
For batching, the number that matters is the water. A batch that is bottled and refrigerated has to carry its dilution with it, because there is no ice in the bottle to supply it. Batch the water in, chill hard, and pour into a chilled glass. If you intend to pour the batch over fresh ice instead, take the dilution figure down — the ice in the glass will supply the rest, and a batch diluted twice tastes exactly as flat as that sounds.
Finished strength of classic specs
| Drink | Spec | Method | In the glass (mL) | Finished ABV | Standard drinks |
|---|---|---|---|---|---|
| Martini | 2.5 oz gin 47%, 0.5 oz dry vermouth 18% | Stirred, 22% | 108.2 | 34.56% | 2.11 |
| Old Fashioned | 2 oz bourbon 45%, 0.25 oz syrup | Built, 15% | 76.5 | 34.78% | 1.50 |
| Manhattan | 2 oz rye 45%, 1 oz sweet vermouth 16% | Stirred, 22% | 108.2 | 28.96% | 1.77 |
| Negroni | 1 oz each gin 44%, Campari 24%, sweet vermouth 16% | Stirred, 22% | 108.2 | 22.95% | 1.40 |
| Margarita | 2 oz tequila 40%, 1 oz triple sec 40%, 1 oz lime, 0.25 oz agave | Shaken, 25% | 157.1 | 22.59% | 2.00 |
| Daiquiri | 2 oz rum 40%, 1 oz lime, 0.75 oz syrup | Shaken, 25% | 138.6 | 17.07% | 1.33 |
| Whisky highball | 2 oz whisky 40%, 4 oz soda | Built, 15% | 204.1 | 11.59% | 1.33 |
The Daiquiri and the highball carry identical alcohol — both start from a 2 oz pour at 40% — yet one is served at 17% and the other at 12%. Strength in the glass and alcohol consumed are different questions.
Where the dilution percentages come from, and how to get your own
The 25%, 22% and 15% figures in the method selector are planning defaults drawn from craft-bar practice, not measurements of your bar. Real dilution depends on ice temperature, ice size, how long you shake, how hard, and how cold the tin was to begin with — a shake with wet ice out of a warm well can add half again what a shake with hard, dry, cold ice adds.
Measuring your own takes one minute and a kitchen scale. Weigh the empty glass. Build and shake the drink exactly as you serve it, strain, and weigh the glass again. Subtract the ingredient weights you poured — or simply pour the same spec without ice and weigh that — and the difference is the melt water in grams, which is millilitres to within a percent at these densities. Do it three times and use the average. Then set the method selector to Enter my own dilution and put that number in.
What throws a cocktail ABV calculation off
- Forgetting the dilution entirely. The single biggest error. A spec that reads 28% in the tin reaches the drinker at around 23% after a normal shake, and at a lower figure still by the last sip.
- Using the bottle's stated ABV for a liqueur. Orange liqueurs range from about 15% to 40% depending on the brand, and vermouths from 15% to 18%. The category tells you very little; read the label.
- Counting free-poured volumes as jigger volumes. A free pour is a habit, not a measurement. If the spec matters, measure it.
- Ignoring what happens in the glass. A drink served over ice keeps diluting. The calculator gives you the strength at the moment it is served, which is the highest it will ever be.
- Batching the dilution twice. Water in the bottle plus fresh ice in the glass is double dilution. Pick one, or split the difference deliberately.
- Confusing strength with dose. A weaker drink in a bigger glass can contain more alcohol. The standard-drinks output is the number to watch, and it does not move when you change the dilution.
- Entering proof in the ABV field. Eighty proof is 40% ABV, not 80%. Both strength fields have a proof option in the unit selector for exactly this reason.
Key terms
- Dilution
- Water added to a drink by contact with ice, expressed here as a percentage of the volume that was poured. A 25% dilution on 120 mL of ingredients means 30 mL of water and 150 mL in the glass.
- Standard drink
- A fixed quantity of pure ethanol used for health guidance. The US definition is 14 g, which is 17.74 mL. The UK unit is 10 mL of ethanol and Australia uses 10 g, so the same cocktail scores differently in each system.
- Batching
- Pre-mixing a cocktail in bulk, including its water, so it can be poured to order. The reason the water has to be measured rather than melted in.
- Spirit-forward
- A drink built only from spirits, fortified wines and bitters, with no juice, dairy or soda. These finish strongest because the only non-alcoholic component is the melt water.
- Overproof
- A spirit bottled above the usual 40% mark, commonly 50% to 75.5% ABV. A small overproof float can move a finished drink by several points of ABV.
Where this sits among the other drinks calculations
This page answers the strength question for a single spec. Two neighbouring questions come up constantly and are handled elsewhere.
The first is quantity: how much to buy for a party or a wedding, which is driven by guest count, event length and drink mix rather than by strength. That is the wedding alcohol quantity calculator, and for draught beer the pour-count arithmetic is in the keg servings calculator. Both take the strength figure from this page as an input to the responsible-service side of the plan rather than computing it.
The second is the strength of the ingredients themselves. If you are making your own, the alcohol comes from fermentation and the arithmetic runs off gravity readings instead of labels: the beer ABV calculator for a wort, the Brix to alcohol calculator for juice, must or honey water. And if you are cutting a high-proof spirit to a house strength for a bar programme, the proof and dilution calculator solves the same ethanol balance in reverse: it tells you the water needed to hit a target strength rather than the strength that follows from a given water addition.
One last observation about the balance itself. Every calculation on this page is the same conservation statement seen from different angles — the ethanol you poured is the ethanol in the glass. Dilution moves the strength, the glass size, and how the drink reads on the palate, and it moves the standard-drink count by exactly nothing. That is worth keeping in mind whenever a spec is adjusted to make a drink feel lighter.