JLPT N4 Study Guide: The Real Test Isn't Grammar, It's Production (2026)

"I would do ok on JLPT and can understand what's happening, but can't actually hold a conversation. The minute I try to respond I just freeze up and my brain ceases to function."

— RiftGazing, on the Bunpro community forums

JLPT N4 is the level where most self-studiers first realize something is wrong. The certificate is in reach. The vocab decks are shrinking. Grammar quizzes feel manageable. And then somebody actually speaks Japanese to you, and the system you spent six months building locks up.

This guide gives you the standard JLPT N4 requirements — vocab, kanji, grammar, study hours, exam format — and then names the trap that the standard guides miss. N4 contains two parallel tests inside it. The certificate scores one. Real life scores the other. Optimize for the wrong one and you get a credential that doesn't match your ability.

What JLPT N4 Actually Measures (The Numbers)#

The 2026 N4 requirements are consistent across every official and unofficial source:

• Vocabulary: ~1,500 words total (your ~800 from N5 plus ~700 new)
• Kanji: ~300 total (your ~100 from N5 plus ~200 new)
• Grammar: ~100 patterns
• Study hours: 300–400 total, typically 3–6 months after N5 at 30–60 minutes daily
• Exam: 115 minutes, three sections
  1. Language Knowledge (vocabulary) — 30 min
  2. Language Knowledge (grammar) + Reading — 60 min
  3. Listening — 35 min

To pass, you must clear both an overall minimum and a sectional minimum. Listening often catches learners who optimized for written grammar: a strong overall score still fails the test if the listening section alone falls below threshold.

The official JLPT guidebook describes N4 as "ability to understand basic Japanese." That phrasing buries the real demand: at N4, you read short passages on familiar topics, follow slow everyday conversations, and parse stacked grammar layers like 〜てから, 〜のに, 〜ば, 〜てもらう under time pressure. It is a real step up from N5, not an incremental one.

The N4 Trap: Two Parallel Tests Inside One Certificate#

Pull two learners who both passed N4 last month. Ask them to introduce themselves and explain why they study Japanese. One does it smoothly with mistakes you'd expect at their level. The other freezes. Same certificate. Opposite real-world ability.

JLPT measures recognition only. There is no speaking section. There is no writing-from-blank section. Every question shows you Japanese and asks you to identify, parse, or comprehend it. You can pass N4 with a perfectly trained recognition system and a completely untrained production system. The certificate cannot tell.

This is the production gap at its sharpest. At N5 the gap is small because the material is small. At N4 it widens fast, because N4 grammar stacks: conditionals, te-form extensions, giving/receiving, register-shifted synonyms. Recognition lets you decode these layers when you see them. Production requires you to assemble them on the fly, which is a different cognitive skill.

The standard N4 guides drill the recognition system and never name the gap. That's why you finish their study plans, pass the test, and still cannot speak. They got you to the certificate. They didn't get you to the language.

The Five Mistakes N5 Graduates Make at N4#

Every one of these is a recognition-system shortcut that breaks at N4 because the material no longer allows it.

Mistake 1: Drilling te-form in isolation#

At N5 you learned how 食べる becomes 食べて. At N4, te-form is no longer the destination. It's the connector. 〜てから, 〜ている, 〜てある, 〜てしまう, 〜ておく, 〜てみる, 〜てくれる, 〜てもらう, 〜てあげる all live on top of it. If your te-form lives only in conjugation drills, the extensions won't stick under conversation speed. Practice te-form inside full sentences that use each extension naturally.

Mistake 2: Memorizing the four conditionals as rules#

たら, ば, と, なら all translate as "if." The textbook rule tables get you through the test. They don't get you through real choice moments. Native speakers pick by feel based on what kind of conditional relationship they mean (temporal sequence, hypothetical, automatic result, contextual response). Memorize the table and you'll freeze deciding which one fits when you actually need to speak.

Mistake 3: Treating transitive/intransitive pairs as vocabulary#

開ける/開く. 閉める/閉まる. 入れる/入る. 出す/出る. The pairs aren't synonyms. They encode whether you're talking about an agent doing something to an object, or the object's resulting state. N4 reading passages and listening clips lean on this distinction constantly. Drilling them as flashcard pairs is fine; using them in sentences where the perspective changes is what makes the difference stick.

Mistake 4: Skipping giving/receiving until later#

あげる, くれる, もらう (and their て-form extensions) are not "extra" N4 grammar. They are how Japanese encodes social direction in everyday speech. Putting them off until "after I pass" means every casual conversation you try will have a perspective gap. Native speakers build entire courtesy into these verbs. Skip them and your Japanese sounds odd even when each word is right.

Mistake 5: Studying kanji standalone#

300 kanji in two months is unsustainable as a flashcard grind. Learn each kanji embedded in 2–3 common N4 words you'd actually use. 仕 alone is abstract; 仕事 (work) and 仕方 (way of doing) make it stick. The kanji is the vocabulary's skeleton, not its own subject.

N4's Iconic Grammar Layers, In Sentences#

These four cards illustrate the production challenges that stall most N5 graduates. Each one shows a grammar layer that recognition study under-prepares you for.

The 4-conditional challenge (〜ば here)#

〜ば signals a clean cause-effect: do X, get Y. Swap to 予約したら and the meaning shifts slightly (more "when you reserve, it'll be cheaper"). Swap to 予約すると and you imply automatic consequence. Same English translation. Three different production choices. The test recognizes any; conversation rewards the right one.

Resulting states via 〜てある#

〜てある says someone cooked the rice and the rice is now in the resulting cooked state. Compare 炊いています (someone is cooking the rice right now) and 炊きました (the rice was cooked, no comment on current state). The transitivity (you cooked it deliberately) and the persistence (it's still in that state) are both encoded. N4 reading passages assume you feel this difference.

Giving/receiving perspective#

〜てくれる locks the perspective: someone did something for me, and I'm the beneficiary. The same English ("My colleague helped me") could in theory drop to 同僚が仕事を手伝いました, but that strips the gratitude and the social direction. In real Japanese it sounds cold. Picking 〜てくれる, 〜てもらう, or 〜てあげる correctly is the difference between speaking Japanese and speaking translated English.

Reason connectors (〜ので vs から)#

から (which you know from N5) and ので both mean "because." ので feels more objective and softer, common in polite contexts. から is more direct and personal, fine among friends but blunter with strangers. N4 listening sections love testing the register match: a worker explaining lateness to their boss should pick ので; a friend explaining flakiness can use から. The pattern this teaches: many N4 "synonyms" carry social register, not just meaning.

A Production-First Study Protocol#

Standard N4 plans tell you to work through Genki II or Quartet, drill Anki, and practice past papers. That stack is the right tool for the recognition half of the certificate; keep it. The production half needs its own training on top, and it requires changing how you spend your daily study time.

The split that works at N4:

• 40% recognition input. Textbook chapters, grammar explanations, kanji decks, listening clips.
• 40% production drill. Build N4 sentences from English prompts. Pick the conditional, pick the giving/receiving verb, pick the reason connector, get feedback on each.
• 20% past-paper conditioning. Time-pressure practice on real JLPT-style questions to keep test stamina sharp.

The 40% production block is what most N5 graduates skip. It's the block that closes the gap.

Studying by yourself at this stage is harder than at N5 because the feedback is harder. You can self-grade an N5 vocab card by checking the answer; you cannot self-grade whether your 〜てくれる was the right perspective choice for the situation. You need either a tutor or a tool that catches the social-register and perspective errors on every sentence you produce.

This is exactly the loop JIVX runs at N4. The prompt says "My colleague helped me with the task," you type 同僚が仕事を手伝ってくれました, and the AI grades the perspective particle in addition to the conjugation. Drop 〜くれました for plain 〜ました and the feedback flags the missing social direction. Pick 〜あげました by mistake and the feedback names the perspective flip. Same sentence, three production attempts, three calibrations to your actual speaking instinct. Recognition study cannot do this because recognition study never asked you to choose.

What "Passing N4" Actually Means#

Honesty about the certificate matters because the disappointment of passing-and-still-freezing is what makes most learners quit somewhere between N4 and N3.

Passing N4 proves you can:

• Recognize ~1,500 N4 words in context
• Read ~300 kanji in everyday Japanese contexts
• Parse stacked grammar (te-form extensions, conditionals, giving/receiving) under time pressure
• Follow slow, clear spoken Japanese on familiar topics

Passing N4 does not prove you can:

• Hold a basic spoken conversation in real time
• Write a paragraph from scratch about your day
• Recall vocabulary in production direction (English → Japanese under pressure)
• Choose between near-synonym grammar (〜ので vs から, たら vs ば) by feel

Both halves are real Japanese skills. The certificate measures one. The other is yours to build deliberately, and the timeline to that level is shorter when you train production alongside recognition rather than after it.

If you can already pass N4 on a practice paper but still freeze in conversation, the gap isn't more study hours. It's the wrong study mix. Add the production block and the freeze starts to go away within weeks.

When a specific N4 pattern keeps tripping you (which conditional fits here, which giving verb, transitive or intransitive), the free grammar reference of 170+ JLPT patterns covers each one with example sentences across N5 through N1.

Frequently Asked Questions#

The short version: the JLPT N4 certificate is real and worth pursuing. It also leaves out the half of Japanese that actually lets you talk to people. Train both halves together. Recognition for the test; production for the language. The certificate doesn't distinguish, but you will know which one you built.