How do you pronounce cà phê sữa đá?

Roughly kah fay, SUH-uh, DAH. Four syllables, each given about the same weight, because Vietnamese does not stress one syllable of a word over another the way English does. The marks are tones: cà falls from mid, phê stays level, sữa breaks and rises, đá rises. No English respelling carries the tones, so treat the spelling above as a way to be understood rather than a way to be correct.

What does cà phê sữa đá mean?

Coffee, milk, ice, in that order. Cà phê is coffee, sữa is milk and means sweetened condensed milk unless someone says otherwise, and đá is ice. Vietnamese drink names are built by stacking these words, so cà phê đen đá is coffee + black + ice and cà phê sữa nóng is the same drink as cà phê sữa đá served hot.

What does phin mean in Vietnamese?

The coffee filter: a small cup with a perforated base that sits on the rim of a glass and lets hot water drip through the grounds. A Vietnamese–Vietnamese dictionary gives the word a second, unrelated sense as well, a thin fine-woven cloth used for shirts and handkerchiefs, and marks the coffee sense as the common one now. Coffee brewed in one is cà phê phin.

How do you pronounce phin?

Feen. Vietnamese has no letter f and spells that sound ph, which is why Kirby's description of Hanoi Vietnamese transcribes the initial of pha, 'to brew', as [f] and the rime of tin, 'news', as [in]. Put the two together and phin is [fin]. It does not rhyme with the English pin and it is not said as two syllables.

Does phin come from the French word filtre?

That is the usual account and we have not found a lexicographic source for it, so treat it as the common story rather than a settled etymology. The shape of the word is at least what Vietnamese would do with filtre: a Vietnamese syllable takes one onset consonant and allows only /p t k m n ŋ j w/ at the end, so a cluster and a final -tre have nowhere to go. That is consistent with the French account without being evidence for it.

Is cà phê a French word?

It is a French borrowing that is now an ordinary Vietnamese word. Trần (2023), surveying Indo-European loanwords in 18th- and 19th-century Vietnamese dictionaries, records the French café entering as cà phe, with the addition of tones among the adaptations it underwent. It is written as two syllables because Vietnamese writes every syllable separately.

Why do the letters have marks over them?

Most of them are tones, and they change which word you have said. Kirby describes Hanoi Vietnamese as distinguishing six tones in syllables that end in a vowel or a nasal: ma is 'ghost' with no mark, mà 'but', má 'cheek', mạ 'rice seedling', mả 'tomb' and mã 'code'. Other marks are part of the letter rather than a tone, which is why ă, â, ê, ô, ơ and ư are their own vowels and đ is its own consonant.

How to Say Cà Phê Sữa Đá, and Phin

The short answer

Say it roughly kah fay, SUH-uh, DAH, with every syllable given the same weight. Each piece is a word: cà phê is coffee, borrowed from the French café; sữa is milk, meaning condensed milk unless someone says otherwise; đá is ice. Phin, the filter, is said feen, because Vietnamese spells the f sound as ph and has no letter f. The marks above the vowels are tones rather than stress, and they change the word: in Hanoi Vietnamese the single syllable ma is six different words depending on which mark it carries.

Cà phê sữa đá, syllable by syllable

The name lists what is in the glass, in order.

Written	Say it roughly	Means
cà phê	kah fay	coffee
sữa	SUH-uh	milk, meaning condensed milk
đá	dah	ice

Three words, four syllables, and all four get the same weight. English picks one syllable and leans on it; Vietnamese does not, so leaning on phê is what makes an English speaker hard to follow, along with flattening the tones.

The respellings on this page are an approximation for English speakers, not a transcription. They get you understood. They do not get you right, because no English spelling can carry a tone.

The marks are tones, and they change the word

Kirby’s description of Hanoi Vietnamese sets out six tones for syllables that end in a vowel or a nasal, and the article makes the point with one syllable and six words:

Written	Tone	Kirby’s gloss
ma	ngang, level	ghost
mà	huyền, mid falling	but, yet
má	sắc, rising	cheek
mạ	nặng, low glottalised	rice seedling
mả	hỏi, low falling	tomb
mã	ngã, broken	code

So the mark on đá is the rising tone. It carries as much of the word as a consonant does.

Syllables that end in a hard stop take a reduced set: Kirby gives two tones there rather than six. That is why the tone system looks like six tones in some descriptions and eight in others, and both are right about different syllables.

Not every mark is a tone. The hook on ơ, the breve on ă, the circumflex on ê and the bar through đ belong to the letter. Vietnamese d and đ are two different consonants: Kirby transcribes d as [z] in da, ‘skin’, and đ as an implosive voiced stop. An English speaker reading đá as “da” has said a different word.

The drinks are built by stacking

Word	Means
cà phê	coffee
sữa	milk, condensed unless stated
đen	black
đá	ice
nóng	hot
trứng	egg
dừa	coconut
muối	salt
chồn	civet

Stack them and the menu reads itself. Cà phê đen đá is coffee + black + ice. Cà phê sữa nóng is the same drink as cà phê sữa đá, served hot. Cà phê trứng is egg coffee and cà phê chồn is civet coffee, which is where the English “weasel coffee” comes from.

Bạc xỉu breaks the pattern, coming from Cantonese rather than Vietnamese; its own page covers what it means. In Hanoi, nâu, brown, stands in for sữa, so the drink there is nâu đá, and the Hanoi and Saigon vocabularies diverge further than that.

What phin means

A Vietnamese–Vietnamese dictionary gives the word two noun senses. The first is the one you came for: a small utensil with a perforated base that sits on the rim of a cup, through which hot water drips over coffee grounds. Coffee made in one is cà phê phin, and the liquid that comes out is nước phin.

The second sense has nothing to do with coffee. Vải phin is a thin, fine-woven cloth, used for shirts, blouses and handkerchiefs. The dictionary marks the coffee sense as the common one today and the cloth as the rarer.

Say it feen, one syllable. Vietnamese has no letter f; the sound is spelled ph, which is why Kirby’s own example word for [f] is pha, ‘to brew’. The rime -in he transcribes [in], in tin, ‘news’. Put them together and phin is [fin]. What the phin physically is, and which one to buy, are on how to use a phin and how to choose a phin filter.

The filtre question

The usual account has phin coming from the French filtre. It is repeated almost everywhere, including elsewhere on this site, and we have not found a lexicographic source for it. It stays an account.

What can be said is that phin is the shape French filtre would take on arrival. A Vietnamese syllable, in Kirby’s description, has one onset consonant, an obligatory vowel and a tone, and licenses only eight segments at the end: /p t k m n ŋ j w/. A fl-type cluster has nowhere to sit and a final -tre has nowhere to go. Reduce filtre to one syllable under those constraints, with /f/ at the front, and phin is what is left.

That is consistent with the French story rather than evidence for it. Any source word beginning with /f/ would come out the same way.

Cà phê came from café

Trần (2023) surveyed the Indo-European loanwords recorded in Vietnamese dictionaries of the eighteenth and nineteenth centuries and lists the French café, read [ka.fe], entering as cà phe. Adding tones is named there as one of the standard adaptations a European word underwent on the way in.

It is written as two words because Vietnamese writes every syllable separately, so a borrowing of two syllables arrives as two.

The borrowing is old enough to be invisible. Cà phê is not a French word used in Vietnamese the way “espresso” is an Italian word used in English. It is the ordinary Vietnamese word, and everything built on top of it — cà phê phin, cà phê sữa đá, cà phê cóc for a pavement stall — is Vietnamese construction on a borrowed root.

When the tone comes out wrong

It usually does not matter.

Nguyen (2019), citing Brunelle (2009), reports that almost all Southern Vietnamese speakers do not distinguish hỏi from ngã in running speech, where Northern speakers do, and that Southern listeners rely on context to tell which of the two was meant. That is a contrast the language itself lets context carry. It does not follow that any tone is optional. It does mean a listener in a coffee shop has a short list of things you might have been trying to say.

The written names are the fallback worth knowing: they stack, and sữa spelled correctly is unambiguous in a way that sua said quickly is not.

Questions people ask

How do you pronounce cà phê sữa đá?: Roughly kah fay, SUH-uh, DAH. Four syllables, each given about the same weight, because Vietnamese does not stress one syllable of a word over another the way English does. The marks are tones: cà falls from mid, phê stays level, sữa breaks and rises, đá rises. No English respelling carries the tones, so treat the spelling above as a way to be understood rather than a way to be correct.
What does cà phê sữa đá mean?: Coffee, milk, ice, in that order. Cà phê is coffee, sữa is milk and means sweetened condensed milk unless someone says otherwise, and đá is ice. Vietnamese drink names are built by stacking these words, so cà phê đen đá is coffee + black + ice and cà phê sữa nóng is the same drink as cà phê sữa đá served hot.
What does phin mean in Vietnamese?: The coffee filter: a small cup with a perforated base that sits on the rim of a glass and lets hot water drip through the grounds. A Vietnamese–Vietnamese dictionary gives the word a second, unrelated sense as well, a thin fine-woven cloth used for shirts and handkerchiefs, and marks the coffee sense as the common one now. Coffee brewed in one is cà phê phin.
How do you pronounce phin?: Feen. Vietnamese has no letter f and spells that sound ph, which is why Kirby's description of Hanoi Vietnamese transcribes the initial of pha, 'to brew', as [f] and the rime of tin, 'news', as [in]. Put the two together and phin is [fin]. It does not rhyme with the English pin and it is not said as two syllables.
Does phin come from the French word filtre?: That is the usual account and we have not found a lexicographic source for it, so treat it as the common story rather than a settled etymology. The shape of the word is at least what Vietnamese would do with filtre: a Vietnamese syllable takes one onset consonant and allows only /p t k m n ŋ j w/ at the end, so a cluster and a final -tre have nowhere to go. That is consistent with the French account without being evidence for it.
Is cà phê a French word?: It is a French borrowing that is now an ordinary Vietnamese word. Trần (2023), surveying Indo-European loanwords in 18th- and 19th-century Vietnamese dictionaries, records the French café entering as cà phe, with the addition of tones among the adaptations it underwent. It is written as two syllables because Vietnamese writes every syllable separately.
Why do the letters have marks over them?: Most of them are tones, and they change which word you have said. Kirby describes Hanoi Vietnamese as distinguishing six tones in syllables that end in a vowel or a nasal: ma is 'ghost' with no mark, mà 'but', má 'cheek', mạ 'rice seedling', mả 'tomb' and mã 'code'. Other marks are part of the letter rather than a tone, which is why ă, â, ê, ô, ơ and ư are their own vowels and đ is its own consonant.

Sources

Figures on this page come from the following. Where a claim is an estimate or a common description rather than a measurement, the text says so.

Vietnamese (Hanoi Vietnamese)Journal of the International Phonetic Association 41(3), 381–392 (James P. Kirby) · 2011 · accessed 2026-10-08Hanoi Vietnamese distinguishing a six-tone paradigm in open or sonorant-final syllables plus a two-tone paradigm in syllables ending in an unreleased oral stop; the tone names ngang, huyền, sắc, nặng, hỏi and ngã with the example set ma, mà, má, mạ, mả and mã; the initial spelled ph transcribed [f] with the article's own example word pha, 'to brew'; the initial spelled d transcribed [z] in da, 'skin'; đ as an implosive voiced plosive; the rime spelled -in transcribed [in] in tin, 'news'; a syllable requiring an obligatory onset, vowel and tone; and the eight segments licensed in coda position, /p t k m n ŋ j w/.
Indo-European Loanwords in Vietnamese Dictionaries in 18th and 19th CenturiesBritish Journal of English Linguistics 11(3), 56–65 (Trần Hương Thục) · 2023-10 · accessed 2026-10-08The French café, read [ka.fe], recorded in the surveyed dictionaries as the Vietnamese cà phe, and the addition of tones listed as one of the adaptations French loanwords underwent on entering Vietnamese.
"Do You Hear What I Hear?": On the Elusive Nature of Vietnamese TonesProceedings of the National Conference on Undergraduate Research 2019, 972–979 (Mai Chi Nguyen) · 2019-04 · accessed 2026-10-08Vietnamese described as having six tones; and the report, citing Brunelle (2009), that most Southern speakers do not make a clear distinction between hỏi and ngã, that almost all of them do not distinguish the two in running speech where Northern speakers do, and that Southern speakers must usually rely on context to tell which is meant.
Tone perception in Northern and Southern Vietnamese (Journal of Phonetics 37(1), 79–96), as listed in the author's own publicationsMarc Brunelle, University of Ottawa · 2009 · accessed 2026-10-08The citation details of the study Nguyen (2019) reports for the Southern hỏi/ngã near-merger, verified against the author's own publication list. We reached the finding through Nguyen (2019) and have not read this paper.
phin — Từ điển Việt ViệtVDict · accessed 2026-10-08A Vietnamese–Vietnamese dictionary giving phin two noun senses: a small utensil with a perforated base that sits on the rim of a cup and through which hot water drips over coffee grounds, and a thin, fine-woven cloth used for shirts, blouses and handkerchiefs (vải phin); the phrases cà phê phin and nước phin; and a note that the coffee sense is the common one today while the cloth sense is used less often.