5 things that help with cold foam that comes out flat and runny instead of thick and stable
Independently chosen — nobody pays to be on a list, and we say what didn't make it. How we pick the 5.
The five, at a glance
1Start at under 4°C, not just 'fridge cold'2Switch to 2% milk, not whole3Add a whisper of heavy cream to the cold 2% milk4Add a very small pinch of salt before frothing5Froth in two stages: fast first, slow to finishStart at under 4°C, not just 'fridge cold'
Most people think 'cold' means whatever comes out of the fridge, which typically runs at 6–8°C — warmer than you expect because of door-opening and thermostat drift. That difference matters more than people realise. Cold foam is a physical structure that starts collapsing the moment it warms. Starting at 2–4°C gives you more working time before friction heat and air incorporation degrade the foam. The cold also firms up fat globules, which reduces how much they interfere with protein bonding during those first critical seconds of agitation. Pre-chilling the frothing vessel is part of this: a warm container pulls heat from the milk the instant you pour.
Switch to 2% milk, not whole
Full-fat milk feels like the logical choice — richer milk, richer foam — but the fat is working against you here. Fat coats the proteins that need to bond together to trap air, and when there's too much of it, those bonds never fully form. Two-percent hits a sweet spot where there's enough protein structure to build a lattice that holds, without the fat smothering it. Skim milk technically foams even stiffer, but it tastes brittle and hollow. Two-percent stays silky and holds its shape for the two or three minutes you actually need. Most specialty cafes use it for cold foam even when whole milk dominates their hot drinks menu — not for flavour reasons, but structural ones.
Add a whisper of heavy cream to the cold 2% milk
This sounds like a contradiction after everything said about fat, but the mechanism is different. A tiny amount of heavy cream — about one part cream to five parts 2% milk — does not undermine foam the way whole milk's fat does. At that concentration, the cream's much higher fat content (35%+) changes the physics: it coats the air-water interface in a way that reinforces bubble walls rather than breaking them down. It's the same principle behind whipped cream — you can't whip skimmed milk into stiff peaks, but pure cream at the right temperature whips fine. The trick is entirely in the ratio. Too much cream and you're back to fat interference; the right amount gives you foam that holds for five minutes instead of ninety seconds.
Add a very small pinch of salt before frothing
Salt does two things to protein structure in cold milk: it slightly alters the electrostatic charges on whey proteins, making them more willing to unfold and bond at low temperatures, and it marginally raises the surface tension of the liquid, which makes bubble walls more resistant to popping. The amount is genuinely tiny — you are not seasoning the foam, you are adjusting the physics of it. This sounds like a trick someone made up, but protein-salt interaction is the same mechanism behind brining poultry or salting pasta water. It's just applied at a much smaller scale.
Froth in two stages: fast first, slow to finish
Most people froth at one speed until it looks done. What this misses is that the two phases of foam-building require different agitation. The first phase — air incorporation — needs fast, vigorous movement to punch air into the milk. The second phase — stabilisation — needs slower, gentler movement to knit that air into a uniform microfoam. Blasting it at full speed the entire time produces foam full of large, unstable bubbles that pop the moment they hit your drink. Slowing down at the end compresses those bubbles into smaller ones that hold. Most people stop 30 seconds too early anyway, when the foam looks done but has not yet reached its final tight structure.
What didn't make the list
This circulates on social media periodically and it does not work. The ice dilutes the milk as it melts, which lowers the protein concentration and makes the foam structure weaker, not stronger. The coldness is real but you can achieve that more reliably — and without interference — by using milk straight from the back of the fridge and pre-chilling your vessel. The ice also physically gets in the way of the frother head. It is one of those ideas that sounds logical until you try it and end up with watery, airy nothing.
Every coffee account recommends it, and commercially it works because barista-edition oat milks contain added gums and stabilisers that ordinary carton oat milk does not have. Standard oat milk from the shelf produces foam that looks promising for about 45 seconds and then reverts to liquid with an almost alarming completeness. Unless you are specifically buying a labelled barista edition, oat milk cold foam at home is a frustration loop — not a dairy-free solution.
Questions people ask
Almost always a temperature problem. If your drink is hot or even warm, the foam is sitting on a thermal updraft that breaks down the bubble walls almost immediately. Cold foam is designed to go on iced drinks — it sits on cold liquid and holds for several minutes. If you pour it onto a hot drink, you are warming it from below and there is no version of cold foam that survives prolonged contact with a 70°C drink. If you want foam on a hot coffee, you want steamed milk foam, which is stabilised through heat rather than cold. Test this by spooning your foam onto an iced drink first — if it holds there, the problem is heat contrast, not foam quality.
Both matter, but milk type and temperature account for more than most people expect. A cheap handheld battery frother with the right cold milk at the right fat percentage will produce better foam than an expensive electric frother used with warm whole milk. That said, a handheld wand frother gives you the most control over depth and speed, which matters for the two-stage approach. Jar shakers that you pump up and down tend to produce large, unstable bubbles because they introduce too much air too fast without the smoothing phase. Electric frother pitchers that heat the milk are the wrong tool entirely for cold foam — ignore the cold setting on them; it rarely gets the milk cold enough to actually work.
Yes, but only reliably for about four hours in a sealed container in the fridge, and the window is shorter than that if the foam was not well-made to begin with. It will settle slightly but a quick ten-second re-froth or vigorous stir can bring it back. Beyond four hours the protein structure degrades and it becomes liquid again regardless of how well it was made originally. If you want to prep ahead, it is faster to keep the milk and cream measured out and very cold, and froth fresh — once the technique is right it takes under a minute.