Your JLPT level is not the number of kanji you own or the months you have studied. It is the point where comprehension stops. Read the six sentences below in order, from pre-N5 up to N1. The last one you understood without translating word by word is your working level. The first one that felt like fog is where your study should start tomorrow.
There is a particular kind of embarrassment that comes with this question. Someone asks what level you are, and you hear yourself say "somewhere around N4, I think," with no idea whether that is true. You have studied for months. You have a flashcard app with an intimidating number in it. And you still cannot answer a one word question about your own ability.
The reason is that most of the things learners count do not measure comprehension. Kanji count measures how many cards you have seen. Study time measures how many hours you sat down. Textbook chapters measure how far a publisher decided to sequence things. None of them tells you whether a real Japanese sentence lands in your head as meaning, which is the only thing the exam actually rewards.
So here is a different approach, and it takes about four minutes. Rather than counting anything, you climb a ladder of sentences until you fall off.
How to use the ladder honestly
Three rules, and the whole thing collapses without them.
- Read the Japanese line first. Do not glance at the hiragana or the English underneath until you have committed to an answer in your head.
- Give yourself one pass. If you need to read it three times to assemble the meaning, that is a fail. The exam gives you roughly a minute per reading question, and re-reading is exactly what runs the clock out.
- Meaning, not translation. If you correctly reconstructed the sentence by translating each chunk into English and gluing them together, mark it as a partial. It counts, but it will not survive exam speed.
The ladder
Want a sharper answer than six sentences?
The free JLPT level test runs twenty five questions, five per level, across kanji readings, vocabulary, grammar and meaning. It shows your accuracy for each level separately, so you see exactly where the drop happens, and every answer comes with an explanation. No signup, nothing stored.
Take the free JLPT level testReading your result
| You stopped at | Your working level | The one thing to fix next |
|---|---|---|
| Pre-N5 | Absolute beginner | Kana until it is invisible, then the first 100 kanji by meaning. |
| N5 | N5 | Verb forms. N4 is the level where one verb turns into a dozen shapes. |
| N4 | N4 | Vocabulary volume. N3 roughly doubles the load and nothing else substitutes for it. |
| N3 | N3 | Reading speed. You know enough words. You cannot yet read them fast enough. |
| N2 | N2 | Formal set phrases and irregular readings, which are recognition items rather than logic. |
| N1 | N1 | Maintenance and output. Rare vocabulary decays faster than anything else you have learned. |
Three things that override the ladder
1. Your listening is probably one level lower
Almost every self taught learner reads above their listening level, often by a full band, because reading lets you set the pace and listening does not. This matters more than it sounds. The JLPT requires a minimum score in each section, not just a good total, so a strong reader with weak ears fails on a technicality. If you cleared N3 above but cannot follow N4 audio at natural speed, your real exam risk sits at N4.
2. Recognition is not recall
This is the exact mechanism behind the famous N3 plateau. Learners keep adding new material on top of a layer that was never properly consolidated, and progress quietly stops looking like progress. The fix is not more cards. It is meeting the same items again on the day you are about to forget them, which is what spaced repetition exists to schedule.
3. The gaps are not evenly sized
N5 to N4 is a short step. N4 to N3 is a long one. N3 to N2 is the longest in the whole exam, because it is where vocabulary roughly doubles and reading shifts from concrete to abstract. If the ladder put you at N3 and the December exam is your target, the honest planning question is not "can I learn N2 grammar in time." It is "can I get my reading speed up in time," which is a completely different training problem.
Which level should you actually register for?
Take the level you can nearly pass today, not the level you would like to be. There are two reasons, and neither is about modesty.
The first is that a JLPT certificate is scored pass or fail, so a near miss at N2 is worth exactly as much on paper as never having sat it. The second is motivational. A deadline you can meet pulls you forward for four months. A deadline you cannot meet turns into a slow, quiet decision to stop opening the app, which is how most people actually quit Japanese. If the ladder put you at N4, register N4 and spend the run up living in N3 material. You will walk into the exam over prepared, which is the only comfortable way to sit one.
From a level to a plan
Knowing your level only helps if something turns it into daily work, and this is the gap where most learners lose the thread. You get a number, you feel briefly clear, and a week later you are back to studying whatever appeared in front of you.
Kanjijo is built to close that gap. You pick your level and get the complete kanji, vocabulary and grammar bank for it, sequenced rather than dumped. Every kanji and every word carries an exclusive mnemonic written to be remembered instead of translated, and every example sentence carries furigana and a full English translation, so you are never guessing at a reading.
The part that changes outcomes is what happens between study sessions. A home screen widget and a lock screen widget keep today's items where your eyes already go, and the test widget quizzes you straight from the home screen without opening anything. The SRS engine schedules each item for the day you are about to lose it, so review stays small instead of piling into a backlog you eventually abandon. When you meet Japanese out in the world, the OCR scanner reads the sign, the menu or the manga page and turns it into study material on the spot.
And because a self check is not an exam, the app carries the rest: JLPT reading sets, JLPT listening sets and full mock tests to confirm the level, plus conversation, shadowing and dictation for the half of the language no reading ladder can measure.
Go deeper
- Free JLPT level test: 25 questions, instant result
- JLPT levels explained: pass marks, structure and exam dates
- The grammar that actually matters at each level
- Why so many learners plateau at N3
- Recognition versus recall, and why it decides your reading speed
- The listening section, and why it fails strong readers
Study the level you actually are
Kanjijo gives you the full kanji, vocabulary and grammar bank for every JLPT level, exclusive mnemonics, furigana on every example, an SRS engine, lock screen and home screen widgets, OCR scanning, JLPT reading and listening practice, mock exams, conversation, shadowing and dictation in one calm app.
Download Kanjijo FreeFrequently Asked Questions
Test upward until you break. Read one representative sentence per level and mark the last one you understood without translating chunk by chunk. That is your working level. Kanji counts and study months measure input rather than comprehension, which is why they mislead so consistently.
N1 is the hardest and N5 is the easiest. The numbering counts downward as difficulty rises: N5, N4, N3, N2, N1. N5 asks for roughly 100 kanji and 800 words. N1 asks for around 2,000 kanji, 10,000 words and the ability to read abstract writing at speed.
The one you can nearly pass today. A certificate is pass or fail, so a near miss at a higher level is worth the same on paper as not sitting it. Register at your current level and spend the months before the exam living in the material one level up.
Because knowing a pattern and reading it at speed are different skills. You have about a minute per reading question, so every word you stop to decode costs you a question somewhere else. The usual cause is recognition without recall: you know the word in a list and cannot retrieve it inside a moving sentence.
Working estimates are about 100 for N5, 300 for N4, 650 for N3, 1,000 for N2 and 2,000 for N1. The exam body does not publish an official kanji list, so treat these as coverage targets rather than pass conditions.