JLPT Listening Tips (N5–N1)
In the JLPT listening section (聴解) every recording is played only once — there is no going back. Most lost points don't come from understanding nothing; they come from not knowing which sentence each item type is really testing: the answer sits in the line after a correction, while the wrong options are things the speakers really did say. This page goes through the techniques item type by item type: what to look at before the audio, what to listen for, the usual traps, and how to use notes and the gaps between items. At the end: how to practise with this site's listening questions and mock tests.
Item Types by Level
| Item type | What it tests | Levels | What the booklet shows |
|---|---|---|---|
| 課題理解 (task-based comprehension) | what someone will do, buy or bring next | N5–N1 | mostly pictures at N5/N4; mostly text from N3 |
| ポイント理解 (point comprehension) | a reason, which one, when | N5–N1 | mostly pictures at N5/N4; text from N3 |
| 概要理解 (summary comprehension) | the main point, the speaker's opinion | N3–N1 | nothing |
| 発話表現 (verbal expressions) | what to say in the pictured situation | N5–N3 | one picture, an arrow pointing at the speaker |
| 即時応答 (quick response) | the most natural reply to one line | N5–N1 | nothing |
| 統合理解 (integrated comprehension) | a judgement combining several speakers or an explanation | N2–N1 | nothing for item 1; later items may print options |
Listening time: N5 30 min, N4 35, N3 40, N2 50, N1 55 (it varies slightly with the recordings). Listening is scored on its own out of 60, and below 19 is a fail even with a passing total (see the December JLPT exam guide).
General Rules
- It plays once — don't replay the last item in your head. If you missed it, guess immediately, mark it, and turn your attention to the next one. The seconds you spend regretting the last item are exactly when the next item's setting is read out.
- Catch who is speaking and who the question is about. Each item opens with something like 「男の人と女の人が話しています」 and the question says 「女の人は…」. If it asks what the woman will do, the man's plans are the distractors.
- What comes after a turn word matters most. The line after 「でも」, 「ただ」, 「実は」, 「〜けど」, 「やっぱり」 or 「それより」 is more likely to hold the answer than the line before it.
- An option isn't right just because you heard it. Distractors are almost always words that really came up in the dialogue. The test isn't "did I hear it?" but "was it rejected in the end, and is it the point the question asks about?"
- You may take notes. You can write on the question booklet. Note only key words, numbers and an "×" next to rejected candidates — never whole sentences.
課題理解: The Last Decision Wins
Before the audio: glance at the options (pictures at N5/N4, text from N3) and see what kind of thing they are — four objects, four times, four combinations of actions — so you have a frame ready.
What to listen for: the question comes first, e.g. 「女の人はこれから何をしますか」, and is repeated after the dialogue. The dialogue usually mentions three or four candidates and then rules them out one by one. The usual ways they're ruled out:
| How it's ruled out | Typical line |
|---|---|
| Already done / already have it | 「それはもう買いました」「まだたくさんあります」 |
| Someone else will do it | 「お茶は私が持っていきます」 |
| Put off until later | 「それは来週でいいです」 |
| Change of plan | 「あ、やっぱりそっちを先にお願い」 |
Technique: when the question says 「これから」, 「まず」 or 「最初に」, you want the first thing to be done next, not everything on the list. From N3 some items need one more step: the agreed time, how long the trip takes, then 「少し早めに出たほうがいい」 — the answer is the time you work out, not the one said aloud. N5 likes numbers and order (which page, which question, what time, how many); write each number down the moment you hear it.
ポイント理解: You Know the Question — Listen for the Reason
Before the audio: you hear the question before the dialogue (and again after it). From N3 the options are printed as text, and at N2/N1 there is time to read them before the dialogue starts. Use it to spot how the four options differ — often only in one point.
What to listen for: the classic questions are 「どうして〜のですか」 (why) and 「何が一番〜ですか」 (what is most …). The most common pattern is a surface reason that gets denied: the other person guesses a reason, the speaker says no, then gives the real one:
- 「忙しかったんですか。」→「いえ、そうじゃなくて、実は…」
- 「値段が高いからですか。」→「それもあるけど、一番は…」
When you hear 「そうじゃなくて」, 「というより」, 「一番は」 or 「実は」, the next sentence is the answer.
Trap: all four options usually do come up in the dialogue — they're just not the point being asked about. Ask "why didn't they go?" and one option will be what they'll do once they get there; ask "what worries them most?" and one option will be a problem already solved. In the N5/N4 "which person?" items (どの人ですか), people are told apart by hat, glasses or where they stand; cross out everyone who doesn't fit as each detail comes.
概要理解: Listen for the Whole, Not the Details
Before the audio: the booklet shows nothing, and the question comes only at the end. You don't know what will be asked, so you have to listen to the whole.
What to listen for: usually one speaker (an announcement, a lecture, a shop assistant explaining something) or two or three people talking, followed by 「何について話していますか」 (what are they talking about) or 「〜はどう考えていますか」 (what does … think). What to catch:
- The beginning and the end — the topic and the conclusion usually sit there;
- The speaker's attitude — for or against, satisfied or not. With a concede-then-turn line like 「〜とは思いますが」 or 「確かに〜。でも…」, the second half is the real opinion.
Traps: options that cover only part of the talk, options that turn a detail into the topic, and options with the attitude reversed. Note only "topic + attitude (○/×)", then mark each of the four options as it is read out.
発話表現: Check the Arrow, Think About Who Speaks (N5–N3)
Before the audio: look at the picture first — who the arrow points to, where they are (a shop, an office, someone's front door), and what they want to do.
What to listen for: the recording describes the situation, asks 「何と言いますか」, and reads three replies (not printed). All three are things you might hear in that kind of place; the two wrong ones are:
- Roles reversed: the person should ask permission, but the option gives it (「どうぞ使ってください」 vs 「使ってもいいですか」);
- Wrong moment: a leaving phrase on arrival (「いってきます」 vs 「ただいま」);
- The other person's line: only the shop assistant says 「いらっしゃいませ」.
Technique: after looking at the picture, think of what you would say if you were the person the arrow points to, then compare with the options as you hear them.
即時応答: Nothing Printed, Decide in Seconds
Before the audio: there's nothing to look at. The booklet has only the answer row — the options are not printed, and you hear them once.
What to listen for: one line, then three replies. While the first line plays, work out its intent — is it asking, requesting, inviting, apologising, complaining? Get the intent right and usually only one of the three replies still fits.
Traps:
- Repeated key words: a wrong option reuses a word from the line, so it sounds related but doesn't answer it;
- Literal vs real meaning: 「時間ありますか」 asks if you're free, not what time it is; 「ちょっと…」 is often a polite no;
- Wrong tense or viewpoint: asked "can you come tomorrow?", the option says "I went yesterday";
- Set phrases at the wrong moment (very common at N5/N4): 「お先に失礼します」→「お疲れさまでした」; you don't answer 「いただきます」 with 「ごちそうさまでした」;
- N2/N1 casual speech: 「〜っけ」, 「〜てもらえない?」, 「〜しっぱなし」, 「〜といて」 — hear the relationship (boss, colleague, friend) and the tone.
Technique: mark each reply ○/×/△ in your head as it plays and fill in the answer as soon as the third one ends. The answering time between short items is only a few seconds — don't wait.
統合理解: Take Notes in a Grid (N2・N1)
Item 1: three people discuss something and reach a decision; the question asks what was decided. No options are printed and the question comes only at the end. The usual pattern: one person proposes, a second objects, and what actually makes the decision-maker choose is someone else's point. So listen hard for the decision-maker's last line: 「じゃあ、〜にします」, 「それなら〜かな」.
The later items: first an explanation (a shop assistant describing four products, someone outlining four plans), then two people discussing which one each will pick, with a separate question about each person's choice. These need notes more than any other item:
- During the explanation, jot a small grid — one or two key words per option (price, time, feature);
- During the discussion, write "M/F" next to each person's conditions, e.g. "F: cheap ×, close ○";
- Then match against the grid. The two people may pick the same option or different ones — don't assume the second answer isn't A because the first one was.
Using the Gaps Between Items
In the real test, the longer items (課題理解, ポイント理解, 概要理解, 統合理解) are followed by roughly a dozen-plus seconds to answer; the short ones (発話表現, 即時応答) by only a few.
- Long items: fill in the answer first, then use the remaining seconds to look ahead at the next item's options or picture — far more useful than agonising over the last one;
- Short items: fill in the answer when the third reply ends, then get ready for the next line;
- Between parts there's a worked example and instructions; if you already know the format, use that time to preview the options of the next few items;
- Fill in the answer sheet as you go, not in the booklet first with the plan to transfer later — there's no extra time once the listening ends.
How to Practise
- Practise one item type at a time: in your level's listening practice, stick to one type until you can say which sentence that type is really testing. After each item you can read the script (with furigana) and the translation and replay it line by line — loop the line you missed.
- Review every mistake: ask yourself which it was — I didn't understand it, I understood but fell for a distractor, or I got the person or the point of the question wrong? The last two are fixed by technique alone.
- Dictate the key line: write down the sentence that holds the answer while replaying it, especially the line after a turn word.
- Do mock tests under real timing: the mock tests play the whole listening section straight through, once, with answering time after each item, just like the exam. Do at least one or two before the test.
| Level | Listening practice | Mock test |
|---|---|---|
| N5 | N5 listening practice | N5 mock test |
| N4 | N4 listening practice | N4 mock test |
| N3 | N3 listening practice | N3 mock test |
| N2 | N2 listening practice | N2 mock test |
| N1 | N1 listening practice | N1 mock test |
All levels start from JLPT listening practice. The site's listening questions are original, written to the official item formats — they are not past papers. For everyday intensive listening there's also the Japanese Daily News podcast.
Related Pages
- December 2026 JLPT exam guide
- Guides: confusable grammar comparisons
- Level overviews: N5, N4, N3, N2, N1
Comments