Why cold brew is quoted as two ratios, not one
Cold brew is almost always made at one strength and drunk at another. You steep at a concentrate ratio — commonly 1:4 to 1:8 by mass — strain, and then cut the concentrate with water, milk or ice at the moment of service. That means two numbers govern the finished drink: the steep ratio, which sets how much coffee is in the vessel, and the dilution ratio, which sets how far that goes.
Brewing as a concentrate has practical reasons behind it. Cold water dissolves coffee solids slowly and selectively, so a long steep at high dose density is the only way to get a full-bodied result without heat. Concentrate also stores compactly and lets a café dial a drink to a customer's taste at the bar rather than committing the whole batch. The cost is that no one can read a cold brew recipe and know how strong the finished drink is without doing both pieces of arithmetic.
The chemistry differs from hot brewing in a way that matters for how you read the numbers. At room temperature the extraction yield — the share of each gram of coffee that dissolves — is lower than the 18–22% a hot Golden Cup brew reaches. Fewer acids and less of the bitter fraction come out, which is the source of cold brew's characteristic smooth, low-acid profile. The practical consequence: a cold brew carrying the same grams of coffee per litre as a filter coffee will not taste as strong, so cold brew recipes run at higher dose density on purpose.
The two-stage formula, and where the missing litre goes
Stage one is simple division. The dose is the batch water divided by the steep ratio: d = W ÷ R. At 1:5 with 2,000 g of water, that is 400 g of coffee.
Stage two is where the arithmetic gets honest. When you lift the filter bag out, the grounds are saturated, and that liquid does not become concentrate. Coffee grounds hold roughly 2 g of water per gram of dry coffee under gravity drainage; squeezing or pressing the bag recovers some of it, at the price of pulling fines and heavier extractives into the batch. So the recovered concentrate is C = W − k·d. Substituting d = W/R gives the compact form C = d(R − k).
That subtraction is brutal at low ratios and mild at high ones. At 1:4 with k = 2, half your water stays in the grounds. At 1:8, a quarter does. At 1:16 — ready-to-drink territory — one eighth does. This is the single biggest reason a cold brew batch under-delivers against the number on the recipe card, and it is why commercial operations either press the bed or brew at higher ratios and accept a longer steep.
Stage three is the dilution. Adding D parts of liquid per part of concentrate multiplies the volume by (1 + D), so the finished volume is V = C(1 + D) and the ready-to-drink ratio — millilitres of beverage per gram of coffee — is (R − k)(1 + D). Notice that the retention appears in the final ratio too. It is not a rounding error you can wave away at the end; it propagates all the way to the cup.
Worked example: a 1 L batch at 1:5, diluted 1:2
You are steeping 1,000 g of water at a 1:5 concentrate ratio, straining under gravity with a retention of 2 g per gram, and serving one part concentrate to two parts water in 240 mL glasses.
- Dose. d = 1,000 ÷ 5 = 200 g of coarsely ground coffee.
- Retained water. 2 × 200 = 400 g stays in the grounds.
- Concentrate recovered. C = 1,000 − 400 = 600 mL. You lose 40% of the batch water to the bed.
- Check by the compact form. C = 200 × (5 − 2) = 600 mL. Same answer.
- Dilution water. 600 × 2 = 1,200 mL.
- Finished volume. 600 + 1,200 = 1,800 mL.
- Drinks. 1,800 ÷ 240 = 7.5 glasses.
- Ready-to-drink ratio. 1,800 ÷ 200 = 9 mL of beverage per gram of coffee, or 1:9 — which is 1000 ÷ 9 = 111 g of coffee per litre of finished drink.
That 111 g/L is nearly double the 55 g/L hot-brew midpoint, and it is not a mistake. Cold extraction pulls out a smaller share of each gram, so the higher dose density lands at a drinkable strength. If you want a lighter drink from the same batch, raise the dilution to 1:3 and the finished volume becomes 600 × 4 = 2,400 mL at 1:12, or 83 g of coffee per litre.
Choosing a steep ratio and reading the result
Pick the steep ratio from how you intend to serve, then let the dilution fine-tune. A 1:4 or 1:5 concentrate is the café standard because it stores in the smallest space and takes heavy dilution with milk without vanishing. A 1:8 concentrate is easier on a home strainer and needs only about one part of water at service. Anything from 1:12 upward is not a concentrate at all — it is cold brew you drink as it comes, and the retention penalty is small enough to ignore.
Judge the result by the ready-to-drink ratio, not the steep ratio. That number is directly comparable between batches and between brewing methods, and it is what actually lands in the glass. A finished ratio between about 1:9 and 1:13 covers most commercial cold brew; below 1:8 the drink reads as syrupy over ice, and above 1:15 it usually reads as thin because cold extraction has not pulled enough out of the grounds to carry that much water.
Two mechanical details move the answer more than people expect. Grind size should be coarse — cold brew runs for hours, and fine grounds both over-extract at the surface and clog the filter, cutting your recovered yield. Ice is dilution. A glass filled with ice and then topped with cold brew can end up half water by the time you finish it, so count the ice as part of D if you serve that way.
Steep ratio reference: dose, concentrate recovered and dilution
| Steep ratio | Dose per litre of water | Concentrate recovered | Water lost to grounds | Dilution for a 1:12 drink |
|---|---|---|---|---|
| 1:4 | 250 g | 500 mL | 50% | 1 : 5 |
| 1:5 | 200 g | 600 mL | 40% | 1 : 3 |
| 1:6 | 167 g | 667 mL | 33% | 1 : 2 |
| 1:8 | 125 g | 750 mL | 25% | 1 : 1 |
| 1:10 | 100 g | 800 mL | 20% | 1 : 0.5 |
| 1:14 | 71 g | 857 mL | 14% | serve neat |
The dilution column comes from (R − 2)(1 + D) = 12. At 1:6, for example, (6 − 2)(1 + 2) = 12.
Cold brew is a refrigeration problem, not just a flavour one
Cold brew never gets hot, so nothing in the process kills microorganisms, and coffee is far less acidic than the pH 4.6 threshold that keeps foods out of the time-and-temperature-control category. Steep in the refrigerator where you can, hold the strained concentrate at or below 5 °C (41 °F), and keep it covered. The FDA Food Code's limit for refrigerated ready-to-eat foods that need temperature control is seven days at 41 °F or below, which is a sensible ceiling to work to even at home.
Sealed or nitro-dispensed cold brew in low-oxygen packaging raises additional questions that home methods do not, and several US state health departments have issued guidance on it. If you are packaging cold brew for sale rather than pouring it fresh, that guidance — not this calculator — governs your process.
Mistakes that ruin a cold brew batch
- Ignoring retention when planning yield. A 1:4 batch gives back only half its water. If a catering order needs 4 L of concentrate, you must steep about 8 L at that ratio.
- Grinding too fine. Fine grounds clog the filter, hold more water, and extract harshly over a long steep. Coarse — like the setting for a French press or coarser.
- Confusing the steep ratio with the drink ratio. 1:5 in the vessel is not 1:5 in the glass. Only the ready-to-drink figure is comparable across recipes.
- Squeezing the bag hard to recover more. It works, but it drags fines and bitter extractives into the concentrate. Press gently or brew at a higher ratio instead.
- Steeping at room temperature for two days. Beyond about 24 hours you gain little and take on a food-safety question. Refrigerated steeps run slower and can justify the longer end of the range.
- Forgetting the ice. A glass of ice can be 40% of its volume in water once melted, which quietly moves your dilution ratio.
- Using the hot-brew Golden Cup band as a target. Cold extraction is lower, so cold brew needs more coffee per litre to taste equivalent. Compare cold batches to each other.
How cold brew relates to the other coffee methods
Every brew method on this site runs on the same mass-ratio arithmetic; what changes is the retention, the contact time and the extraction the water achieves. For hot filter brewing, where a single ratio describes both the vessel and the cup, use the coffee to water ratio calculator. For espresso, where the ratio is beverage mass out over dry dose in and lands between 1:1 and 1:3, use the espresso brew ratio calculator.
Two other approaches sit between them. Japanese iced coffee brews hot directly onto ice, so it keeps hot-brew extraction and aromatics but arrives chilled; the total water — ice included — still uses the hot-brew ratio. Flash-chilled batch brew does the same at café scale. Both taste brighter and more acidic than cold brew, because the acids cold water leaves behind are precisely what heat pulls out.
If you are costing cold brew for a menu, remember that the coffee dose per finished drink is several times a filter coffee's. A 350 mL cold brew at a 1:12 ready-to-drink ratio carries 350 ÷ 12 = about 29 g of coffee, against about 25 g for the same glass of filter coffee brewed at 1:16, which yields roughly 14 mL of beverage per gram once the grounds have taken their share. Push those figures through the recipe cost per serving calculator before you set the price, and check the result against your target with the food cost percentage calculator.