Vietnamese Cold Brew: Robusta at 1:5

Cold brew and iced coffee are different drinks

The line is drawn at the moment of brewing, not the moment of serving. The National Coffee Association puts it plainly: “Iced coffee is coffee that has been brewed hot using a standard method, then cooled down to serve. Cold brew is brewed with room temperature or cooler water and can be served hot or over ice.”

Cà phê sữa đá (Vietnamese iced coffee with condensed milk) is iced coffee by that definition. Hot water goes through a phin in four to six minutes, onto condensed milk, then over ice. Nothing in it is cold brewed.

Vietnamese cold brew is the other method, run on Vietnamese beans. Cold or room-temperature water, coarse grounds, hours instead of minutes. Vietnamese sellers label it cà phê ủ lạnh, coffee steeped cold.

Both end up in a tall glass of ice with condensed milk in them, which is why the two names get swapped. The drinks taste different and they fail differently.

Why robusta suits cold water

Cold water is a poor solvent, so cold brew compensates with time and dose. What it gets more of and less of depends on what is in the bean.

A 2021 study in LWT ran arabica and robusta through identical cold-brew conditions, two grind sizes, 5 °C and 15 °C, infusions up to 25 hours. Its own summary of the result: “Cold brews from Robusta coffee had higher caffeine and CGA contents and higher AA”, meaning chlorogenic acids and antioxidant activity.

That gap is the same one robusta shows in hot brewing. Measured mean caffeine in green beans is 2.9% by dry weight for robusta against 1.8% for arabica, and chlorogenic acid runs higher in robusta too. Both read as bitterness.

Cold brewing does not take them out. It changes what arrives alongside them, which is the whole of what the method does.

The same study found that coarser particles and lower temperatures extracted less of everything. If a cold brew is coming out thin, grinding a notch finer or leaving it on the counter rather than in the fridge will do more than another six hours.

The roast is the awkward part

Most coffee sold as Vietnamese is dark roasted, and dark roast is the one variable in cold brew that works against you.

Fuller and Rao, cold-brewing at 35 g to 350 mL at room temperature, found 3-chlorogenic acid “at higher concentrations in cold brew coffee made with medium roast coffees, rather than dark roast”, with medium roasts also giving more caffeine than dark, though not always by a significant margin.

So the dark roast that makes a phin work gives up less in cold water than a medium roast of the same bean would. The practical reading is that a medium-dark bag is an easier starting point for cold brew, and a very dark one wants a tighter ratio to land in the same place. That second half is an inference from their results, not something measured on Vietnamese beans. Vietnamese roast styles covers how wide that range runs, because the traditional dark roast is not the only thing Vietnamese roasters make.

The ratio, and what you get for it

Ratio	Per litre of water	What it is
1:4	250 g of coffee	Concentrate, tight. Dilute roughly 1:3 with water or milk
1:5	200 g of coffee	Concentrate. The phin’s ratio, twelve hours instead of six minutes
1:8	125 g of coffee	Closer to ready to drink. Not what the NCA recommends
1:15	65 g of coffee	Filter-coffee strength. It will not survive condensed milk and ice

The National Coffee Association recommends 1:4 to 1:5 on a coarse grind. The two looser rows are there for scale, not as recommendations. We have not brewed these, and the quantities are starting points aimed at the strength the drink needs rather than tested recipes.

At 1:5 a cold brew sits at roughly the concentration a phin produces, which is three times filter strength and the reason the drink survives two tablespoons of condensed milk and a full glass of melting ice.

Twelve hours, and why six might do

Twelve hours is the figure the NCA gives, and the overnight habit it fits, grind at night and strain at breakfast, is most of why it stuck.

The research suggests it is longer than it needs to be for the two compounds anyone has tracked. Fuller and Rao sampled cold brews over time and found that caffeine and 3-chlorogenic acid “reached equilibrium according to first order kinetics between 6 and 7 hours in all cold brew samples instead of 10 to 24 hours outlined in typical cold brew methods”.

Two compounds are not flavour, that work used light and medium roast arabica, and nobody has published the same curve for dark-roast robusta. Twelve hours remains the safe default. Seven is worth trying before assuming a longer steep is earning its time.

Acidity, and what the study does not cover

Cold brew’s reputation for low acidity has one good study behind it and a caveat worth stating.

Rao and Fuller measured pH and total titratable acidity in cold and hot brews of the same coffees. pH came out comparable, 4.96 to 5.13 for cold brew against 4.85 to 5.10 for hot, while the hot brews carried consistently higher titratable acidity. By the measure a pH meter gives, the two are about equally acidic; the cold brew holds fewer of the acids a palate reads as sourness.

The caveat is the coffee they tested: light-to-medium roast arabica from Brazil, Colombia, Ethiopia, Mexico and Myanmar. A dark Vietnamese robusta starts somewhere else entirely, and that comparison has not been published.

Sweetness comes from the tin

Nothing in the method adds sugar. A black cold brew of dark Vietnamese robusta is about as unsweet as coffee gets, which is the opposite of what most people expect from the phrase.

Condensed milk goes in afterwards rather than underneath. Stir two tablespoons into about 60 ml of concentrate until the colour is even, then pour it over ice. Add the milk to the ice first and it sinks and stays there.

A 30 ml pour of sweetened condensed milk carries about 21 g of sugar and about 120 calories. The arithmetic for every Vietnamese drink, and for which tin changes it, is on the calories page.

Caffeine

Cold water gets the caffeine out. Fuller and Rao found substantially more caffeine in their coarse-grind cold brews than in hot brews of the same coffee at the same ratio, with no significant difference at a medium grind.

The dose and the ratio decide the rest. A 1:5 cold brew holds about as much coffee per millilitre as a phin brew, so a glass of undiluted concentrate lands in phin territory, which our caffeine page estimates at 320–490 mg for a 20 g phin against the 113–247 mg the FDA states for a 12 fl oz cup of brewed coffee. Dilute the concentrate three to one and the figure falls with it, while adding ice leaves it where it was.

What cold brew is not good at

It takes twelve hours and a phin takes six minutes, and most of the rest follows from that.

It gives no feedback while it runs. A phin tells you something every single brew, because the drip rate is the one signal the device produces and reading it is how you fix the next one. A jar in the fridge says nothing until you strain it, by which point every decision has already been made.

It does not make cà phê sữa đá. That drink depends on hot coffee meeting condensed milk and dissolving it on contact, and on what a dark roast gives up under hot water in four minutes, commonly described as heavy and faintly scorched. Cold brew with condensed milk over ice is a good drink built from the same ingredients. It is a different one.

It spends robusta’s advantage in the wrong direction. The compounds robusta brings more of are the bitter ones, and cold brewing keeps them. Anyone reaching for cold brew because they want a smoother, sweeter cup has picked an odd bean to do it with. The reason to use this one is that twelve hours of cold water treats a dark robusta differently from six minutes of hot water, not that it sands the bean down.