Scope. This page provides a controlled reader-run protocol and does not claim site-run measurements. For cooking, cooling, storage and reheating, follow the current official guidance linked on the page.
Cold brew coffee is usually explained as a convenience. Put grounds in cold water, wait, strain, and you have coffee without a kettle. The explanation for why it tastes different is normally that it is "smoother", which is not an explanation.
The actual mechanism is more useful, because it generalises. Cold coffee is a different selection of compounds pulled out of the same beans, and once you understand that, the same move works on tea, on aromatics, on dairy and on stock.
Extraction as selection, not just as speed
Here is the principle, stated once, plainly. Everything else on this page is a demonstration of it.
When you steep something, the temperature determines which compounds come out, and the time determines how much of them. Temperature is a selector. Time is a volume control.
Different compounds have different solubilities, and those solubilities respond differently to temperature. Some dissolve readily in cold water. Some barely move until the water is hot. Some are only produced by heat acting on the material in the first place, so no amount of cold steeping will ever produce them, and some are volatile enough that heat drives them off before you get to drink them.
That means a cold steep and a hot steep of the same material are not the same drink at different strengths. They contain different things in different proportions. A cold steep run for a long time can be stronger than a hot one and still taste less bitter and less complex, because the compounds that carry bitterness and much of the complexity were never dissolved.
Two consequences follow, and they are the useful ones:
- If a hot infusion is giving you a quality you do not want, cold steeping can often leave that quality behind rather than dilute it.
- If a hot infusion is giving you a quality you do want, cold steeping will probably not give it to you at any duration, and the answer is to fix the hot method instead.
That second point is why "just do it cold" is not general advice. Cold is a different tool with a different output.
Coffee: what cold leaves behind
Coffee is the clearest case because the difference is large and almost everyone has tasted both sides of it.
The same grounds steeped cold and steeped hot produce noticeably different results, in a consistent direction: the cold version reads as less bitter and less acidic, and it also tends to read as less aromatic and less varied. Whatever compounds carry the sharper edges of a hot cup, they are pulled out much more readily by hot water.
The trade is honest in both directions, and people who prefer hot coffee are not wrong. The complexity in a good hot cup is partly made of the compounds cold water leaves in the grounds. Removing the harshness removes some of the interest with it.
Grind size matters more in cold extraction than in hot, because the process is slow enough that the distance a dissolved compound has to travel out of the particle becomes limiting. That is diffusion doing the work, the mechanism described in when the variable is hours, not degrees.
Times and ratios. We give none. Measure extraction against duration at a fixed grind and ratio, tasting at intervals, and record the curve rather than a single time, because the useful thing is where the curve flattens.
Tea: tannins, time and temperature
Tea shows the principle sharply because a longer or hotter steep can taste more astringent and drying. This article does not assign that perception to one compound class; it compares steep time, temperature and the tasted result for the named tea.
That astringency is temperature-sensitive as well as time-sensitive. Steeped cold, tea does not develop it in the same way over the same period, which is why a cold steep can run for an extraordinarily long time without becoming undrinkable, while a hot steep goes past the point of pleasure in minutes.
This is the practical version of the principle: cold steeping widens the window. A hot steep has a narrow window because the compounds you want and the compounds you do not want are both coming out fast. A cold steep separates them in time as well as in solubility, so you are much harder to punish for leaving it too long.
Which is better depends on what you want from the leaf. A tea whose interest lies in aromatic top notes may lose most of what makes it worth drinking when steeped cold. A tea that is robust and tends toward astringency may be transformed.
Times per tea type. We give none. Measure against the same fixed variables and note the point at which the cold steep stops changing.
Aromatics: cold-steeped versus warmed
This is the application people have not usually thought of, and it is where the idea earns its place in a cooking site rather than a drinks one.
Take an aromatic and put it in a liquid. Citrus zest in a syrup. Vanilla in cream. Spices in vinegar. Herbs in a dressing base. You have two ways to do it: warm the liquid, or leave it cold and wait.
The two produce different results in a predictable direction. Warming pulls compounds out faster and pulls out more of the less-soluble ones, and it also drives off the most volatile aromatics as it goes. Cold steeping is slower, more selective, and keeps the volatile top notes that heat would remove.
So the choice is a real choice, not a convenience:
| You want | Steep |
|---|---|
| The bright, fresh, volatile top of the aromatic | Cold, and long |
| The deeper, heavier, less soluble part | Warm, and shorter |
| Both | Split the batch, do each separately, combine |
That last row is the technique worth taking away. Splitting an infusion and running half of it cold and half warm, then blending, gives you a range of compounds that neither method produces alone. It costs nothing but a second container.
Times per aromatic. We give none, because zest, whole spices, ground spices and fresh herbs all behave differently, and the difference between whole and ground is likely to be larger than the difference between cold and warm, since particle size sets the diffusion distance.
Anything oil-based goes elsewhere. If your liquid is oil rather than water, vinegar or alcohol, stop reading this page. Oil infusions with fresh aromatics carry a specific and serious hazard, the guidance is not something to infer from a general principle, and it is written up with its sources in how to infuse oil, safely. There is no oil guidance on this page and there never will be.
Dairy and stock infusions
Two cooking applications where the principle is already in use, usually without being named.
Dairy. Infusing cream or milk with an aromatic is nearly always taught as a warm process: bring it up, take it off, leave it to steep. That works, and it is also a selection. A cold infusion of the same aromatic in the same dairy gives a different profile, generally lighter and more volatile-forward. Whether you want that depends on what the dairy is going into.
Stock. A stock is an extraction, and the argument of stock below the simmer is that temperature changes what comes out, not just how fast. Held gently, you extract selectively and the liquid stays clear. Boiled, you extract more of everything and drive fat into the liquid, and clarity goes. A cold extraction of stock ingredients is a further step in the same direction, with a limited output, since some of what a stock is for only appears with heat.
Both of these run into the safety question immediately. Anything held for hours at kitchen temperature, and anything involving dairy or animal ingredients held below cooking heat, is governed by rules we do not write. We state no holding temperature, no duration and no storage guidance here. Before setting up any long steep involving dairy, meat or fish, read FDA's safe food handling guidance, checked 27 September 2026 or your own national authority's equivalent; it covers refrigerating perishables and how long they may sit out.
How to convert a hot infusion to a cold one
If you have a hot infusion you like and want to try the cold version, the conversion is not a matter of swapping the temperature and keeping everything else.
- Decide what you are trying to keep or lose. If you are chasing the volatile top notes or escaping astringency, cold is likely to help. If the quality you like is deep, heavy or only produced by heat, cold will not give it to you.
- Expect to increase the material, the time, or both. A cold steep at the same ratio for the same duration will be weaker, because you have slowed the whole process down.
- Reduce particle size, within reason. Diffusion distance is the limiting factor in a slow steep. Grinding, cracking or zesting more finely does more for a cold infusion than it does for a hot one.
- Taste on a schedule rather than at the end. Draw a small sample at intervals and keep the samples. That is how you find where the curve flattens for your material, and it is the only way to arrive at a time that means anything.
- Strain properly and promptly. A cold steep that sits on its solids continues to change, and past a point it stops improving.
- Check the food-safety guidance first if the ingredients are perishable. Coffee, tea and dry spices in water are a different proposition from dairy, meat, fish or fresh produce held for hours.
Conversion factors between hot and cold times. We give no multiplier, because a single multiplier across materials would be an invented number, and the whole point of the selectivity principle is that different materials behave differently.
In Plain English
Steeping temperature decides which compounds you pull out. Steeping time decides how much. Cold extraction is not weak hot extraction; it is a narrower selection, which is why cold coffee tastes less bitter and less complex at the same time, and why cold-steeped tea can sit for hours without turning astringent.
The rule of thumb: cold and long keeps the bright volatile parts and leaves the harsh, heavy parts behind. Warm and short does the opposite. If you want both, split the batch and run it both ways.
This helps you when an infusion is coming out with a quality you do not want, and you have been trying to fix it by adjusting the time. Often the time is not the problem and the temperature is. Use it when you are setting up any steep, including the ones in a recipe that only ever mentions warming.
What to do next: take an infusion you already make warm, split the ingredients, run half cold, and taste them side by side. That single comparison teaches the principle better than the rest of this page. If the liquid is oil, do not run that experiment from this page. Read how to infuse oil, safely first.
Where to go next
Record your own times for each application and plot them as curves rather than collapsing them into single recommended durations.
The technique library has the rest of the extraction and infusion methods, and when the variable is hours, not degrees sets out the four things time does to food that heat cannot, of which this is one.
Keep the worse and null outcomes in the same notes as the apparent improvements.