JLPT N4 vs N3: What Actually Changes
You passed N4, or you are about to, and the question is whether December should be N3. The vocabulary roughly doubles and the test gets 25 minutes longer, but neither of those is what fails people. The scoring does.
The Numbers Side by Side
| N4 | N3 | |
|---|---|---|
| Pass mark (out of 180) | 90 | 95 |
| Total time | 115 min | 140 min |
| Vocabulary booklet | 25 min | 30 min |
| Grammar + Reading booklet | 55 min | 70 min |
| Listening booklet | 35 min | 40 min |
| Scored sections | 2 | 3 |
| Section minimums | 38 / 120 and 19 / 60 | 19 / 60, three times |
| Kanji you are expected to read | about 300 | about 650 |
| Vocabulary | about 1,500 words | about 3,700 words |
Pass marks, times and section minimums are the official figures published on jlpt.jp. Kanji and vocabulary counts are estimates: the JLPT stopped publishing official lists in 2010, and every textbook count differs a little.
The Scoring Change Nobody Warns You About
At N4 your Language Knowledge and Reading answers are added together into one score out of 120, and you need 38 of it. A student who is strong on vocabulary and grammar can carry a weak reading performance inside that combined number and never find out. I have watched it happen for years: the N4 certificate arrives, the student is pleased, and the reading gap is invisible.
At N3 the gap becomes visible. Reading is scored on its own, out of 60, and you need 19. Language Knowledge is scored on its own, out of 60, and you need 19. Listening, the same. Miss any one floor and the total does not matter.
So when you move from N4 to N3, ask yourself one thing before anything about vocabulary: can I get 19 out of 60 on reading, on the clock, with no help from my grammar score?
"I Got Almost 100% on N4. What Would I Score on N3?"
I get asked this more than anything else at this stage, and I am not going to give a number, because any number would be made up. There is no conversion, for two reasons that both come from how the test is built.
First, JLPT scores are scaled. Your 170 on N4 is a position relative to that sitting's difficulty, not a raw count of correct answers, and an N3 paper is scaled against N3 candidates. Second, and more important, the N4 number hides the split. Two students with 170 on N4 can have completely different reading ability, because reading was never reported separately.
What a near-perfect N4 does tell you: your grammar and vocabulary foundation is sound, and your listening is probably fine, because N4 listening is already at natural speed. Those two sections will most likely clear their N3 floors after a normal study cycle. The unknown is reading. Measure it directly: sit a timed N3 reading section and count.
What I actually tell students
Take one N3 reading section, 70 minutes for the whole grammar-and-reading booklet is the real time, so give the reading part about 40. If you finish and get around 60% right, register for N3. If you run out of time with two passages unread, that is your answer too, and it is a better answer than any prediction from your N4 score.
What Is New on the Paper
Most of the N3 paper looks like N4 with harder material: kanji reading, kanji writing, context vocabulary, grammar selection, sentence ordering, short and mid-length passages, information retrieval, and four of the five listening tasks. Two question types are new, and one old one changes its weight.
Long reading passages (ๅ ๅฎน็่งฃใป้ทๆ)
One passage of roughly 550 characters with four questions, alongside the short and mid-length passages you already know from N4.
Why it matters: At N4 you could re-read a whole passage for every question. At 550 characters you cannot. You need to read once, hold the structure, and go back only for detail.
Listening for the gist (ๆฆ่ฆ็่งฃ)
Three questions where nothing is printed on the paper. You hear a short talk, then the question, then four spoken options.
Why it matters: Every other listening task lets you read ahead. This one does not, so if your habit is to lean on the printed choices, this is where it shows.
Reading as its own score
The same reading questions you know from N4, now producing their own 60-point score with a 19-point floor.
Why it matters: Covered above. I list it again because this is where strong N4 candidates usually lose N3 on the first attempt.
Should You Skip N4 and Go Straight to N3?
You can. The JLPT has no ladder rule; nothing stops a first-time candidate registering for N3. In some countries the fee is the same at every level, and in Japan it is ยฅ7,500 whichever you pick, so the cost of a miss is a wasted sitting rather than wasted money. The people I advise to skip are readers. The people I advise not to skip are the ones who have never been timed.
Go straight to N3 if
- โYou can finish an N3 reading section in the time allowed with about 60% correct
- โYou already read Japanese you were not assigned: news, manga, game text, work email
- โN4 grammar is automatic, including the conditionals (ใใ, ใฐ, ใชใ, ใจ) and passive
- โYour N4 result, or a full N4 mock, came back at 150 or more out of 180
Sit N4 first if
- โYou have never sat a timed reading section at any level
- โYour N4 total was fine but you could not say which part was weakest
- โYou still translate sentences in your head to understand them
- โYou want the certificate for a visa or job application this year and cannot afford a miss
Pass Rates in 2025
| Sitting | N4 certified | N3 certified | Examinees (N4 / N3) |
|---|---|---|---|
| July 2025 | 36.8% | 38.0% | 178,849 / 218,726 |
| December 2025 | 34.6% | 31.2% | 194,015 / 249,389 |
Japan and overseas combined. Source: jlpt.jp, July 2025 and December 2025. Examinee counts include people who sat only some sections.
N3 is the most-taken level in the world, ahead of N2, which explains how often I hear this question. The pass rates of N4 and N3 sit close together, and I read that as a statement about the candidates rather than the material: people arriving at N3 have, on average, prepared more than people arriving at N4, and the third scored section takes back roughly what that extra preparation gave them.
Frequently Asked Questions
Can I skip N4 and take N3 directly?
Yes. The JLPT has no prerequisites, so you can register for N3 without ever sitting N4. Whether you should depends on reading: N3 scores Reading as its own section with a minimum of 19 out of 60, and that is the section that fails most people who jump. If you can finish an N3 reading section on time with about 60% correct, skipping N4 is reasonable.
If I nearly aced N4, what score would I get on N3?
There is no conversion. N4 reports one combined score for Language Knowledge and Reading, so a high N4 total says little about your reading by itself, and N3 gives Reading its own pass line. A near-perfect N4 usually means grammar and vocabulary are safe. To find out about reading, sit a timed N3 reading section and count.
Is N3 harder to pass than N4?
The pass mark rises from 90 to 95 out of 180, the test grows from 115 to 140 minutes, and Reading becomes a separately scored section with its own minimum. In 2025 the overall pass rates were close: 36.8% and 34.6% for N4 in July and December, 38.0% and 31.2% for N3 (jlpt.jp). What changes is the shape of the test, not how many people get through.
N3 Study Guide โ
How to prepare for each section
N3 Passing Score โ
Section minimums and the score checker
N5 vs N4 โ
The step before this one
Find out where your reading really is
Nihongo Pass scores every mock exam by section and predicts your pass probability from the section floors, not the total, so a hidden reading gap shows up in practice instead of on results day.