What we measured
We analysed 41,762 graded Hangeul drill attempts from 328 distinct learners on Lingrow between 3 July and 8 September 2026. A recognise attempt means hearing the sound and picking the letter, or reading the letter and picking the sound. A produce attempt means building or writing the letter from its sound.
Good rate is the share rated Good or Easy, ratings 3 or 4 on the 4-point scale. In these drills that means the learner got the answer right first time. Again rate is the share rated Again, rating 1, which means a wrong answer. We included only letters with at least 40 reviews and 30 learners. The course launched in the App Store on 22 July 2026, so attempts before that date come from pre-release testing.
This is data from one app's learners over about two months. It measures performance in Lingrow's graded drills, not every part of learning or using Hangeul, and we'll update the page as more attempts come in.
Hardest Hangeul letters to recognise
The recognition ranking begins with ㅋ and ㅚ. The next group contains several consonants from Korean's aspirated and tense series, alongside the plain consonants ㄷ and ㅂ.
| Rank | Letter | Sound | Type | Graded attempts | Learners | Good rate | Again rate |
|---|---|---|---|---|---|---|---|
| 1 | ㅋ | k | consonant/aspirated | 490 | 55 | 39% | 38% |
| 2 | ㅚ | oe | vowel/compound | 606 | 41 | 39% | 38% |
| 3 | ㅍ | p | consonant/aspirated | 656 | 60 | 43% | 35% |
| 4 | ㄸ | tt | consonant/tense | 330 | 43 | 43% | 37% |
| 5 | ㅊ | ch | consonant/aspirated | 452 | 56 | 44% | 35% |
| 6 | ㄷ | d | consonant/basic | 1,243 | 142 | 44% | 36% |
| 7 | ㅌ | t | consonant/aspirated | 574 | 58 | 44% | 35% |
| 8 | ㅉ | jj | consonant/tense | 386 | 45 | 45% | 36% |
| 9 | ㅕ | yeo | vowel/basic | 680 | 107 | 46% | 36% |
| 10 | ㅂ | b | consonant/basic | 1,256 | 142 | 47% | 36% |
| 11 | ㅃ | pp | consonant/tense | 370 | 52 | 47% | 31% |
| 12 | ㄲ | kk | consonant/tense | 421 | 50 | 50% | 32% |
| 13 | ㅈ | j | consonant/basic | 915 | 144 | 50% | 32% |
| 14 | ㅙ | wae | vowel/compound | 336 | 43 | 51% | 30% |
| 15 | ㅓ | eo | vowel/basic | 1,800 | 321 | 51% | 36% |
| 16 | ㅑ | ya | vowel/basic | 766 | 108 | 51% | 32% |
| 17 | ㅜ | u | vowel/basic | 2,081 | 320 | 52% | 33% |
| 18 | ㄱ | g | consonant/basic | 1,508 | 148 | 54% | 31% |
| 19 | ㅗ | o | vowel/basic | 1,619 | 319 | 54% | 33% |
| 20 | ㅒ | yae | vowel/compound | 358 | 53 | 55% | 27% |
Aspirated and tense consonants dominate the hard end
Six of the eight hardest letters to recognise were aspirated or tense consonants. ㅋ had a 39% good rate and a 38% Again rate. The other aspirated consonants near the top were ㅍ at 43%, ㅊ at 44% and ㅌ at 44% good.
Tense consonants appeared in the same difficult part of the ranking. ㄸ had a 43% good rate, ㅉ had 45%, ㅃ had 47% and ㄲ had 50%.
Korean stops come in three series: plain, aspirated and tense. English has two (b versus p), so an English speaker files three Korean letters under two English sounds. The aspirated series is the closest to the p, t and k at the start of "pin", "tin" and "kin". That's the problem, because English speakers put that puff on the plain letters too.
Plain stops are difficult inside the same contrasts
The plain consonants ㄷ and ㅂ were also near the hard end. ㄷ had a 44% good rate across 1,243 attempts by 142 learners. ㅂ had a 47% good rate across 1,256 attempts by 142 learners.
Their wrong picks show that the issue isn't confined to aspirated or tense letters. Learners picked ㅌ instead of ㄷ 51 times and ㅂ instead of ㄷ 50 times. They picked ㅍ instead of ㅂ 84 times.
Confusion also ran in the other direction. Learners picked ㅂ instead of ㅃ 37 times and ㅂ instead of ㅍ 35 times. They picked ㄷ instead of ㄸ 37 times, ㄱ instead of ㅋ 47 times, and ㅊ instead of ㅈ 65 times.
These results fit the pronunciation problem. A learner has to hear and produce each consonant as one member of a three-way contrast. One romanised sound can't stand in for one Korean letter.
The easiest recognition letters formed a different group
At the easiest end, ㅣ had a 76% good rate and a 17% Again rate across 1,365 attempts by 320 learners. ㄹ had a 75% good rate and a 17% Again rate across 884 attempts by 192 learners.
ㅡ also had a 75% good rate, with a 16% Again rate across 419 attempts by 108 learners. ㅁ had a 74% good rate and an 18% Again rate across 865 attempts by 193 learners.
So "is Hangeul hard" has no single answer. Learners on Lingrow got ㅣ right first time 76% of the time and ㅋ only 39% of the time in the same drills.
Most common wrong picks
Wrong picks reveal which letters learners grouped together. The table covers errors across all the Hangeul drills in the data window.
| Target | Picked instead | Wrong picks |
|---|---|---|
| ㅜ (u) | ㅗ (o) | 257 |
| ㅗ (o) | ㅜ (u) | 210 |
| ㅓ (eo) | ㅗ (o) | 149 |
| ㅕ (yeo) | ㅑ (ya) | 124 |
| ㅓ (eo) | ㅏ (a) | 89 |
| ㅂ (b) | ㅍ (p) | 84 |
| ㅗ (o) | ㅓ (eo) | 76 |
| ㅈ (j) | ㅊ (ch) | 65 |
| ㅓ (eo) | ㅜ (u) | 63 |
| ㅑ (ya) | ㅕ (yeo) | 63 |
| ㅏ (a) | ㅓ (eo) | 59 |
| ㄷ (d) | ㅌ (t) | 51 |
| ㄷ (d) | ㅂ (b) | 50 |
| ㅋ (k) | ㄱ (g) | 47 |
| ㅛ (yo) | ㅠ (yu) | 46 |
| ㅃ (pp) | ㅂ (b) | 37 |
| ㄸ (tt) | ㄷ (d) | 37 |
| ㅍ (p) | ㅂ (b) | 35 |
| ㄹ (r) | ㄴ (n) | 33 |
| ㅙ (wae) | ㅘ (wa) | 30 |
The largest wrong-pick pattern was ㅜ, ㅗ and ㅓ
Learners picked ㅗ instead of ㅜ 257 times. They picked ㅜ instead of ㅗ 210 times. Those were the two most common wrong picks in the dataset.
ㅓ and ㅗ were also repeatedly confused. Learners picked ㅗ instead of ㅓ 149 times and ㅓ instead of ㅗ 76 times. They also picked ㅏ instead of ㅓ 89 times and ㅜ instead of ㅓ 63 times.
The mouth positions explain what learners need to practise. ㅓ is unrounded, with relaxed lips, while ㅗ is rounded, with the lips forward. English speakers say both as "o", which turns 커피 (coffee) into 코피 (nosebleed).
Romanisation labels ㅓ as "eo" and ㅡ as "eu", but those spellings don't tell an English reader what to do with their mouth. The written letter, the Korean audio and the mouth position need to be learnt together.
Hardest Hangeul letters to write from the sound
Producing a letter means building or writing it after hearing its sound. This ranking does not match the recognition ranking.
| Rank | Letter | Sound | Graded attempts | Learners | Good rate | Again rate |
|---|---|---|---|---|---|---|
| 1 | ㅋ | k | 317 | 51 | 50% | 31% |
| 2 | ㄱ | g | 1,191 | 145 | 52% | 34% |
| 3 | ㄷ | d | 729 | 117 | 53% | 32% |
| 4 | ㅚ | oe | 357 | 38 | 53% | 30% |
| 5 | ㅎ | h | 801 | 140 | 54% | 30% |
| 6 | ㅌ | t | 350 | 52 | 55% | 29% |
| 7 | ㄹ | r | 756 | 109 | 56% | 29% |
| 8 | ㅂ | b | 843 | 138 | 57% | 28% |
| 9 | ㅁ | m | 750 | 190 | 57% | 32% |
| 10 | ㅅ | s | 968 | 188 | 57% | 29% |
| 11 | ㅍ | p | 394 | 58 | 58% | 24% |
| 12 | ㅛ | yo | 459 | 60 | 58% | 28% |
| 13 | ㅜ | u | 1,939 | 301 | 58% | 32% |
| 14 | ㄲ | kk | 281 | 50 | 58% | 27% |
| 15 | ㅓ | eo | 1,584 | 152 | 58% | 28% |
Recognising and producing a letter are different skills
ㅋ topped both rankings, sharing first place with ㅚ for recognition. It had a 39% good rate for recognition and a 50% good rate for production. The rest of the order changed substantially.
ㄱ was much nearer the hard end for production. It had a 52% production good rate across 1,191 attempts, compared with a 54% recognition good rate across 1,508 attempts.
ㅎ gives a clearer contrast between the rankings. Its production good rate was 54% across 801 attempts, placing it near the hard end. Its recognition good rate was 70% across 756 attempts.
The reverse happened for ㅃ. It was difficult to recognise, with a 47% good rate and a 31% Again rate. Yet it was the easiest letter to produce, with an 88% good rate and an 8% Again rate across 154 attempts by 51 learners.
This matters because reading a letter, hearing its sound and writing it from memory place different demands on a learner. Practice needs to cover each direction.
What this means for how to learn
Learn the consonant contrasts side by side. Say 가, 카 and 까 as a set, then listen again before repeating them. A hand in front of your mouth can help you check the puff on the aspirated sound.
Record yourself against Korean audio and get feedback that rates the individual sound, not only the whole word. A word recogniser can guess the word from context even when your ㅍ was really a ㅂ. Our Korean pronunciation guide explains the plain, aspirated and tense series, along with the ㅓ and ㅗ mouth positions.
Learn each sound with its Hangeul letter from day one. Avoid relying on romanisation as the main cue, especially for vowels labelled "eo" and "eu". You can hear letters individually in Lingrow's free interactive Hangeul chart.
Then practise recognising, producing and writing each letter separately, because the data shows they are different skills. The Lingrow Hangeul course inside the Lingrow iOS app was built for exactly this. It rates your pronunciation of individual letters and syllables, and its contrast drill checks that 가, 카 and 까 come out as three different sounds. It checks handwriting from memory and brings each letter back with spaced repetition on a per-letter strength grid. It's a subscription with a free trial.
How to cite this data
This dataset comes from Lingrow's own learners and covers graded Hangeul drills from 3 July to 8 September 2026. Suggested citation: Lingrow, "Hardest Hangeul Letters for English Speakers, Measured (2026)", lingrow.io/blog/hardest-hangeul-letters-for-english-speakers, updated 9 September 2026. You're welcome to quote any number on this page with a link back.
This page will be updated as more data arrives. These numbers describe learners on Lingrow during the stated window, not all Korean learners.
Frequently asked questions
Which Korean letters are hardest to learn?
For learners on Lingrow, ㅋ and ㅚ had the lowest recognition good rates, both at 39%. They were followed by ㅍ and ㄸ at 43%, then ㅊ, ㄷ and ㅌ at 44%. This ranking covers recognition drills, while producing letters from sound gave a different order.
What is the difference between ㅗ and ㅓ?
ㅗ is produced with rounded lips, while ㅓ is unrounded and uses relaxed lips. Learners on Lingrow picked ㅓ instead of ㅗ 76 times and picked ㅗ instead of ㅓ 149 times. Practising the mouth position alongside the written letter helps separate the sounds.
Why is ㅋ so hard for English speakers?
ㅋ belongs to Korean's aspirated consonant series. English speakers put that same puff of air on plain ㄱ too, so the two letters blur into one English k. Among learners on Lingrow, ㅋ had a 39% recognition good rate, and ㄱ was picked instead 47 times.
Are Korean vowels or consonants harder?
The Lingrow data doesn't support one answer for every letter. The hardest recognition rows include aspirated and tense consonants, but compound vowel ㅚ also had a 39% good rate. Vowel confusions were especially common: ㅜ was picked as ㅗ 257 times, while ㅗ was picked as ㅜ 210 times.
Is Hangeul hard to learn?
Difficulty varies by letter and skill among learners on Lingrow. Recognition good rates ranged from 39% for ㅋ and ㅚ to 76% for ㅣ. Producing letters from sound followed another pattern, ranging from a 50% good rate for ㅋ to an 88% good rate for ㅃ.
