Chinese tones: the 4 Mandarin tones, explained
Mandarin Chinese has four tones plus a light neutral tone. Each tone is a pitch shape (high and level, rising, dipping, or falling), and changing the shape changes the word: mā (媽, mother), má (麻, hemp), mǎ (馬, horse) and mà (罵, to scold) are four different words.
This page shows what each tone looks like, the mistake learners make most often with it, and how tones change when two of them meet. The examples use Traditional characters, with pinyin and 注音 (Zhuyin), as they are taught in Taiwan.
Looking for one word? See how to say hello, thank you and other common words, each with a recording.
The four tones on one chart
Linguists describe tone height on a scale from 1 (the bottom of your voice) to 5 (the top). This chart is not drawn by hand: each line is the average pitch of many real words recorded by a native Taiwan Mandarin speaker, sampled across the length of the syllable.
Textbooks give each tone a fixed number (first tone = 55, second tone = 35, third tone = 214, fourth tone = 51). Real speech is close to that, but not the same, and FlappyTone's tunnels follow real speech. Each tone below says what the textbook teaches and what the recordings show. Two things stand out on the chart. First, the tones are relative to your own voice, so a deep voice and a high voice draw the same shapes at different pitches. Second, the second and third tones both dip and rise in a careful citation form; what separates them is how low the dip goes and how early the rise starts.
First tone: high and level (ˉ)
Start high in your voice and hold it flat, like singing one long note. Chao numbers write it 55: it starts at 5 and stays at 5. It is not louder than the other tones, only higher and steadier.
Common mistake: Starting at a normal speaking pitch, so it sounds flat but low, or letting it sag at the end as if finishing an English sentence. Keep it up to the last moment.
In real speech
In our native speaker's recordings, first tone sits at about 4.6, between 4 and 5, not at the very top. It stays there at everyday speed too. Aim for high and steady, not for your highest note.
- 媽mother
- 書book
- 花flower
Second tone: rising (ˊ)
Start in the middle of your range and rise to the top, like a surprised “what?” in English. Chao 35. Native speakers often hold the start for a moment and then rise smoothly through the rest of the syllable.
Common mistake: Dipping too deep before the rise, which makes it sound like a third tone, or rising too little, which makes it sound like a first tone. The turn should be early and shallow.
In real speech
A real second tone is not a straight line from 3 to 5. Our speaker starts near 2.8, dips a little to about 2.3 in the first third, then climbs to near 5. At everyday speed it often stops around 4.5.
- 人person
- 錢money
- 學to study
Third tone: low, dipping (ˇ)
Drop to the bottom of your range and stay low. Said alone or at the end of a phrase it comes back up, so the full shape is a dip: chao 214. Inside a phrase it usually stays low and never rises (the “half third tone”).
Common mistake: Treating it as a rising tone with a small dip. The low part is what makes it a third tone; the rise is optional. Many learners also do the full dip-and-rise on every third tone, which sounds slow and exaggerated in normal speech.
In real speech
The textbook 214 ends at 4. In our speaker's careful recordings the dip reaches about 1.6 halfway through and the rise goes almost to 5. At everyday speed the dip is shallower (about 1.9) and the rise stops near 4, and inside a phrase there is often no rise at all.
- 馬horse
- 好good
- 你you
Fourth tone: falling (ˋ)
Start high and fall fast to the bottom, like a firm “no!” or “stop!”. Chao 51. It is usually the shortest of the four tones.
Common mistake: Not falling far enough, so it sounds like a soft first tone, or starting too low, so there is no room to fall. Start at the top and commit to the drop.
In real speech
The textbook 51 starts at the top. Our speaker's fourth tone peaks a little later, at about 4.5, then falls to about 1.4. At everyday speed it falls less, to about 1.9, so the speed of the drop matters more than how far it goes.
- 罵to scold
- 大big
- 快fast
The neutral tone (˙)
Some syllables, mostly grammar words and the second half of some words, are short and light with no shape of their own. Pinyin writes them with no mark; 注音 puts a dot (˙) in front. Their pitch depends on the tone before them: higher after a third tone, lower after a first, second or fourth tone.
Common mistake: Giving a neutral syllable a full tone. 什麼 is “shén me”, not “shén mè”. Keep it short and let it follow the syllable before it.
- 什麼what
- 你們you (plural)
- 包子steamed bun
Tone sandhi: how tones change next to each other
Dictionaries and pinyin always write a word's own tone. In speech, some tones change when another tone follows them. This is called tone sandhi (變調). Four rules cover almost everything a learner meets, and every example below is a real recording: tap “See it” to watch the pitch the speaker actually made.
1. Third tone + third tone → second tone + third tone
When two third tones meet, the first one rises like a second tone. 你好 is written nǐ hǎo and said ní hǎo. In our recording of 可以, the first syllable climbs to about 4.3 before the second one dips to the floor. Studies of everyday Taiwan Mandarin find the changed tone is very close to a real second tone, though not always identical, so saying it as a second tone is the right target.
- 可以can, maysaid ké yǐ
2. Third tone before any other tone stays low (半三聲)
Before a first, second or fourth tone, or a neutral syllable, a third tone drops low and stays there. It never makes the full dip-and-rise you learn for a single syllable. Chao numbers write this “half third tone” as 21 instead of 214. In 喜歡, the first syllable sinks to about 1.5 and holds; the jump up you hear is 歡 starting high and level.
- 喜歡to like
- 美國America
3. 一 (yī, one) changes with the next tone
Alone, at the end of a word, when counting and in ordinals (第一), 一 keeps the first tone. Before a fourth tone it becomes a second tone: 一下 is said yí xià, and the recording rises before it falls. Before first, second and third tones it becomes a fourth tone: 一起 is said yì qǐ, a fall from the top.
- 一下a bitsaid yí xià
- 一起togethersaid yì qǐ
4. 不 (bù, not) becomes a second tone before a fourth tone
不 is a fourth tone, except before another fourth tone, where it becomes a second tone: 不錯 is said bú cuò, 不是 bú shì, 不要 bú yào. Before any other tone it stays bù: 不好 bù hǎo, 不來 bù lái. In our recording of 不錯, 不 rises from about 2 to the top before 錯 falls.
- 不錯not badsaid bú cuò
FlappyTone's tunnels are measured from these recordings, so a tone pair's tunnel already has the change built in. Fly 可以 as ké yǐ and you stay inside it. The game shows the spoken tones above each word, next to the written pinyin.
Taiwan vs mainland Mandarin tones
Taiwan's 國語 and mainland China's 普通話 use the same four tones and the same sandhi rules. The differences are in which words get which tone, and in how some tones are usually said:
- Fewer neutral tones in Taiwan. Many two-syllable words that end light on the mainland keep a full tone in Taiwan: 朋友 péngyǒu, 學生 xuéshēng, 先生 xiānshēng. In our recording of 朋友, 友 makes a full third-tone dip, not a short light syllable.
- Some words have a different standard tone. 星期 (week) is xīngqí in Taiwan and xīngqī on the mainland. 垃圾 (garbage) is lèsè in Taiwan and lājī on the mainland. 和 (and) is often said hàn in Taiwan.
- The third tone is usually low. Taiwan speakers use the low 21 form in most positions, keeping the rise mostly for a word said alone or with emphasis.
- Writing. Taiwan uses Traditional characters and teaches pronunciation with 注音 (Bopomofo); the mainland uses Simplified characters and pinyin.
- 朋友friend
Both are standard, and speakers of each understand the other without trouble. FlappyTone's recordings are Taiwan Mandarin, so the tunnels follow Taiwan's tones and Taiwan's rhythm.
Free tools for practising the tones
- Chinese tones test: hear ten words and swipe each to its tone. No microphone needed.
- Mandarin tone visualizer: say a tone and see your pitch drawn over a native speaker's.
- Pinyin tone marks converter: turn ni3 hao3 into nǐ hǎo and 注音.
- Common words with audio: hello, thank you, sorry and more, each with a native speaker's recording.
Can't hear the tones? Common questions
Test your ear first: the free Chinese tones test plays ten words and shows which tones you mix up.
Why can't I hear the difference between Chinese tones?
Because English uses pitch for emotion and questions, not for word meaning, so your ear has learned to ignore it inside a word. That is a habit, not a lack of talent. Tones become easier to hear once you know what shape to listen for, so start by learning the four shapes, then listen for them one pair at a time.
Am I tone deaf if I can't hear Mandarin tones?
Almost certainly not. Real tone deafness (amusia) is rare. Most learners who struggle with tones can tell a question from a statement in their own language, which uses the same pitch movement. What is missing is practice with pitch as a part of the word.
Which tones are hardest to tell apart?
Second and third tones. Both go down and then up, so the difference is how low the dip goes and when the rise starts: second tone turns early and rises over most of the syllable, third tone drops low first. First and fourth tones are usually the easiest pair.
How do I practise hearing tones?
Listen to a native speaker say minimal sets like mā, má, mǎ, mà, and say what you hear before you check. Then say them yourself and record or watch your own pitch. Seeing the shape of your voice makes it much easier to notice what your ear is not yet catching.
Do I need to learn tones if people understand me from context?
Context helps, but many common words differ only by tone, and wrong tones make you much harder to understand, especially in short answers with little context. Getting tones right early is also easier than fixing a habit later.
How long does it take to get Chinese tones right?
Most learners can produce the four tones in single words within a few weeks of regular practice. Keeping them right in fast, connected speech takes longer. Short, frequent practice out loud works better than long sessions.