LingairaMinimal Pairs Practice: Fix the Sounds People Keep Mishearing
You say “I need to leave at six.” The other person looks puzzled and asks how long you have lived there. Your grammar was fine. Your vocabulary was fine. One vowel was not.
That is a minimal pair at work: two words that differ by exactly one sound and nothing else. Live and leave. Ship and sheep. Bad and bed. When the two come out of your mouth the same way, the listener has to guess — and some of the time they guess wrong.
Minimal pairs are the most efficient pronunciation exercise there is, because they fix the smallest unit that actually breaks understanding. Here is the list that matters and how to drill it.
What a minimal pair actually tells you
Most pronunciation advice is about sounding better. Minimal pairs are about something smaller and far more useful: being understood.
A pair isolates one contrast. If ship and sheep sound the same when you say them, you have not got a bad accent — you have merged exactly one vowel contrast. That is a small, findable, fixable thing. Most learners have four or five of them, not fifty.
So you are never repairing “my English pronunciation”. You are repairing a short list of contrasts your first language never needed.
Your ears go first
Here is the step almost everyone skips.
You cannot reliably produce a contrast you cannot hear. If bad and bed sound identical to you, your mouth has no target to aim at. You will produce whichever one your first language has, and you will not notice you did it.
Every language trains its children to ignore differences that do not matter in that language. That filter is still running years later, and it quietly deletes the English contrasts your language never used. Minimal pairs work because they push the deleted contrast back in front of your ear, again and again, until the filter lets it through.
This is why the drill below starts with listening and only then opens your mouth.
The contrasts worth your time
You do not need all of these. You need the two or three your first language does not distinguish. Read the list out loud and notice which pairs come out identical — those are yours.
/ɪ/ vs /iː/ — short i against long ee ship / sheep · live / leave · bit / beat · fill / feel · sit / seat · it / eat It is not only about length. The short one is looser and more open, with a relaxed jaw; the long one is tense, with the corners of your mouth pulled back.
/æ/ vs /e/ — bad against bed bad / bed · man / men · sat / set · had / head · bat / bet The /æ/ sits lower and lasts longer than most learners expect. Drop your jaw further than feels polite.
/l/ vs /r/ light / right · lock / rock · play / pray · collect / correct · glass / grass The difference is where your tongue goes. For /l/ the tip presses just behind your top teeth. For /r/ it touches nothing at all — it curls back and floats.
/b/ vs /v/ and /p/ vs /f/ berry / very · boat / vote · ban / van · curb / curve pan / fan · pull / full · copy / coffee For /b/ and /p/ the lips meet and the air stops. For /v/ and /f/ your top teeth rest on your bottom lip and the air keeps flowing — you should be able to hold the sound for three seconds.
/v/ vs /w/ vine / wine · vest / west · veil / whale · verse / worse Same teeth-on-lip for /v/. For /w/ nothing touches: the lips round, as if you were starting to whistle.
/θ/ vs /s/ and /t/ — the th think / sink · thin / tin · three / tree · path / pass · mouth / mouse Tongue tip lightly between the teeth, air passing over it. You should feel the air on your tongue, not behind it.
/n/ vs /ŋ/ — sin against sing sin / sing · thin / thing · ban / bang · win / wing · ton / tongue
Final voicing — back against bag back / bag · cap / cab · seat / seed · right / ride · rope / robe This one is bigger than it looks. Here is the trick: English marks the difference mostly on the vowel, not the final consonant. The vowel in bag is noticeably longer than the vowel in back. Stretch the vowel and the consonant takes care of itself.
The drill: ten minutes, four steps
1. Hear it (2 minutes). Play the pair — a dictionary app, a recording, your speaking app — and just listen. Do not repeat yet. Let your ear collect evidence.
2. Test yourself (2 minutes). Have the two words played in random order and decide which one you heard before you look. Getting this to eight out of ten is the real milestone. Until you can hear the difference, drilling your mouth is guesswork.
3. Say it (3 minutes). Now produce them, one at a time, slowly and exaggerated. Record yourself. Play it back and ask one question: could a stranger tell these two apart?
4. Break the pattern (3 minutes). Say them in random order, then mixed into other words. Saying “ship, sheep, ship, sheep” twenty times teaches your mouth a rhythm, not a contrast — and the contrast is the point.
Then put the pair in a sentence
Word lists are the easy part. Real speech is fast, and the sound has to survive being surrounded by other sounds.
- “I need a ship.” / “I need a sheep.”
- “Can you fill it?” / “Can you feel it?”
- “He’s going to leave.” / “He’s going to live.”
- “She sat down.” / “She set it down.”
- “It’s in the back.” / “It’s in the bag.”
Say each sentence and have someone — or something — tell you which one they heard. This is the only test that counts, because it is the test a real conversation runs.
Three ways this goes wrong
Drilling pairs you already say correctly. It feels productive and changes nothing. Spend your ten minutes on the pair that embarrasses you, not the one you are good at.
Never leaving the word list. Words in isolation transfer to real speech surprisingly badly. Move to sentences by day three, and to a real conversation by the end of the week.
Practising with no feedback. This is the fatal one. If you cannot yet hear the difference between ship and sheep, you also cannot hear whether your own ship sounded like a ship. You need something outside your own ear — a recording you compare, a patient friend, or a tool that scores the sound.
A two-week plan for one contrast
- Days 1–3, 10 minutes: listening and self-testing only, plus slow production at the end. One contrast. Not two.
- Days 4–7, 10 minutes: minimal pair sentences, recorded and played back.
- Week 2: pick five words with that sound that you actually use — your job, your city, your daily life — and work them into conversations on purpose.
- End of week 2: record yourself telling a two-minute story. Count how many times the sound slips. That number is your progress, and it is worth more than a feeling.
Then choose your next contrast. Two contrasts fixed properly will do more for you than twenty half-drilled.
Where minimal pairs stop
Minimal pairs fix individual sounds. They do not fix word stress, rhythm, or the melody of a sentence — and those break understanding just as often. A perfectly pronounced word with the stress on the wrong syllable is still a word people miss.
So run minimal pairs alongside the bigger picture: our guide to improving English pronunciation covers stress and connected speech, and the shadowing technique trains the melody that single sounds cannot. If you are practising without a partner, the methods for practising alone will tell you where to fit these ten minutes.
Where Lingaira fits
The hard part of minimal pairs on your own is step 2 and step 4 — finding out whether the difference actually came out of your mouth.
Lingaira scores your pronunciation word by word after every conversation, so you can see which sounds came through clearly and which ones did not, and it has pronunciation drills that focus on a single sound at a time. You speak, you find out, you say it again better — which is the whole loop, running in ten minutes.
Not sure which sounds to start with? Take the free English level test, then let the feedback from your first few conversations pick your contrast for you.
You do not have to fix your accent. You have to stop two words being the same word. Read the list above out loud tonight, find the pair that comes out identical, and give it ten minutes tomorrow.