JLPT guide
How the JLPT is scored
Every level is out of 180 — and the total is not the rule that fails people.
Almost everything confusing about the JLPT result slip comes from one fact: your score is not a percentage of the questions you got right. Two people can answer the same number of questions correctly in two different sittings and get different scores, and neither slip is wrong.
Every level is marked out of 180
N5 and N1 both total 180 points. What changes is how those points are divided, and that division is the part that decides whether you pass.
| Level | Scored sections | Pass mark |
|---|---|---|
| N5 | Language knowledge & Reading (0–120) · Listening (0–60) | 80 / 180 |
| N4 | Language knowledge & Reading (0–120) · Listening (0–60) | 90 / 180 |
| N3 | Language knowledge (0–60) · Reading (0–60) · Listening (0–60) | 95 / 180 |
| N2 | Language knowledge (0–60) · Reading (0–60) · Listening (0–60) | 90 / 180 |
| N1 | Language knowledge (0–60) · Reading (0–60) · Listening (0–60) | 100 / 180 |
Read that table twice, because it contains something people assume is a misprint: N3 has a higher pass mark than N2. 95 against 90. N2 is unquestionably the harder exam — it just does not need as high a score, because the scale is set per level and the two numbers are not measuring the same thing.
The rule that fails people who passed
Clearing the total is not enough. Every scored section also carries a minimum, and missing any one of them fails the whole exam no matter how high the total is.
- N1, N2, N3 — at least 19 out of 60 in each of the three sections.
- N4, N5 — at least 38 out of 120 in Language knowledge & Reading, and 19 out of 60 in Listening.
So an N2 candidate scoring 55, 55 and 18 has 128 points against a pass mark of 90 — comfortably over, and a fail. The 18 in listening ends it. This is the single most common way people are surprised by a result slip, and it is why practising only your strongest section is a bad strategy: a section you have written off can sink an exam you would otherwise pass easily.
It also means the useful question is never "what is my total?" but "what is my weakest section?" Every paper on this site is marked section by section for exactly that reason.
Why the scores are scaled
The JLPT uses scaled scoring — 尺度得点. Raw answers are converted onto a common scale using item response theory, so a mark means the same thing whatever else was in your booklet.
The reason is fairness across sittings. No two papers are exactly equal in difficulty, and the test runs twice a year worldwide. If results were raw percentages, an easy December would hand out more passes than a hard July for no reason connected to the candidates. Scaling removes that.
The practical consequences are worth knowing:
- You cannot calculate your score from a count of correct answers, and nobody can tell you exactly how many you needed.
- Questions are not all worth the same in the final scale, and the weighting is not published.
- Any site promising you an exact JLPT score, including this one, is estimating. A practice result tells you where your weak section is. It does not tell you what the real slip will say.
How long each level takes
| Level | Total test time | Kanji | Vocabulary |
|---|---|---|---|
| N5 | 90 minutes | ~100 | ~800 |
| N4 | 115 minutes | ~300 | ~1,500 |
| N3 | 140 minutes | ~600 | ~3,500 |
| N2 | 155 minutes | ~1,000 | ~6,000 |
| N1 | 165 minutes | ~2,000 | ~10,000 |
The kanji and vocabulary figures are the commonly used estimates. The JLPT has published no official word list since 2010, so treat them as a sense of scale, not a syllabus.
What to check on results day
Look at the section scores before the total. If you failed, the section under its minimum tells you what the next six months are for. If you passed, the lowest section is still the one to work on: the next level up assumes it.