JLPT N1 vocabulary: how to close the 10,000-word gap

N1 is commonly described as requiring around 10,000 cumulative words. If you passed N2, you know roughly 6,000 — which leaves a 4,000-word gap that no single textbook covers, because N1 famously tests "anything an educated adult might read." Word lists alone stop working at this scale. What works is a hybrid: a frequency-ranked core list, reading-driven acquisition for the long tail, and a review system engineered to survive a deck of thousands. This guide covers all three.

1. The gap in numbers

*Coverage is the share of running words in a text you already know. The jump from 95% to 98% sounds small but is the difference between "reading with a dictionary" and "reading" — at 95%, one word in twenty is unknown, roughly one per sentence; at 98% you can usually infer the rest from context. Closing the gap is exactly what buys you that last 3%.

2. Prioritize by frequency, not by list order

The 4,000 new words are not equally likely to appear. Corpus frequency is your best predictor of both exam value and real-life value, so learn in bands:

1. Band 1 — high-frequency N1 core (~1,500 words). Words that appear in almost every editorial and formal document: 措置（そち）"measure," 懸念（けねん）"concern," 促進（そくしん）"promotion/encouragement," 著しい（いちじるしい）"remarkable, striking." Learn these from the structured N1 word list at a fixed daily pace.
2. Band 2 — mid-frequency (~1,500 words). Common in specific genres (politics, economy, science columns). Acquire mostly through reading + mining (section 4), backed by the list.
3. Band 3 — long tail (~1,000 words). Literary verbs, onomatopoeia, four-character idioms (四字熟語) like 一石二鳥（いっせきにちょう）"two birds with one stone." Do not grind these from a list first — let reading select which ones you actually meet, then keep them.

The practical rule: list-first for Band 1, input-first for Bands 2–3. People who try to brute-force all 4,000 from flashcards alone usually stall around month 4 under review debt.

3. Kanji compounds (漢語) vs native words (和語)

Most of the N1 gap consists of Sino-Japanese compounds (漢語, kango) — two-kanji, on-reading words used in formal writing. Their everyday native (和語, wago) equivalents you already know. N1 loves testing the formal member of the pair:

This structure is a gift: kango are combinatorial. Once you know 削減（さくげん）"reduction" and 費用（ひよう）"expenses," then 費用削減 "cost cutting" is free. Study kanji readings systematically and your effective vocabulary compounds — literally. When you meet a new two-kanji word, always ask: do I know each character's on-reading from another word? Link it in your notes if so.

4. Reading-driven acquisition: mining NHK and beyond

From Band 2 onward, your textbook is the news. A workable daily routine (30–40 minutes):

1. Pick one NHK news article per day (regular NHK NEWS WEB, not the Easy version — you're past that). Politics and economy articles have the highest N1-word density.
2. First pass: read for meaning without stopping. Mark unknown words, don't look them up yet.
3. Second pass: look up the marked words, but only keep the ones that pass the mining filter below. A typical article yields 5–10 unknowns; keep 3–5.
4. Add keepers as sentence cards — the NHK sentence itself, with reading and meaning — into your review rotation on the review deck.

The mining filter: keep a word if (a) you've seen it twice in a week, (b) it's built from kanji you know, or (c) it's clearly a genre workhorse. Skip proper nouns and one-off technical jargon. Typical high-value NHK catches: 政権（せいけん）"administration/government," 見通し（みとおし）"outlook/forecast," 相次ぐ（あいつぐ）"to occur one after another" — as in 値上げが相次いでいる（ねあげがあいついでいる） = "price hikes are coming one after another," a sentence pattern NHK uses almost weekly.

Round out the diet with one longer-form source you enjoy — a novel, a column, long-form interviews. The N1 reading section includes essays and opinion pieces whose tone is closer to a paperback than a news wire.

5. Synonym clusters: the real N1 boss fight

N1 vocabulary questions rarely ask "what does X mean?" They ask "which of these four near-synonyms fits this sentence?" So stop learning words in isolation and start learning them in clusters with contrast notes:

Build one cluster per study day out of words you've already met individually (the dictionary is useful for pulling members together). Write one sentence per member where the other members would sound wrong. That "sounds wrong" instinct is precisely the skill the exam measures — and it's also what makes your own Japanese sound adult.

Other clusters worth building early, because they recur across past papers: the "gradually" family — 次第に（しだいに, formal written）, 徐々に（じょじょに, steady measurable change）, だんだん (conversational); the "excel" family — 優れる（すぐれる）, 秀でる（ひいでる）, 抜群（ばつぐん）; and the "vague" family — 曖昧（あいまい）, 漠然（ばくぜん）, 不明瞭（ふめいりょう）. Three or four clusters a week for six months covers most of what Part 1 of the vocabulary section can throw at you.

6. Retention at scale: keeping 10,000 words alive

The hardest part of N1 vocabulary isn't learning — it's not forgetting words #1–6,000 while adding #6,001–10,000. Concrete tactics:

• Do the math on review load first. 4,000 new words over 12 months ≈ 11 new words/day, which generates roughly 100–150 due reviews/day at steady state. Budget 30–40 minutes daily for reviews before any new material. If reviews exceed 200/day, pause new cards until the queue drains.
• Use graduated intervals, not mass rereading. A Leitner-style system (like Mirai Voca's wrong-answer deck: miss → box 0, correct answers promote, box 3 graduates) concentrates effort exactly on the words that need it.
• Kill leeches deliberately. A word failed 5+ times needs a new memory hook, not more repetitions: write a personal example sentence, tie it to a kanji you know, or park it in a synonym cluster. Repetition without change just burns time.
• Keep old levels warm through input, not cards. Your N2 and lower vocabulary gets maintained for free by the daily NHK reading — that's another reason the reading habit is non-negotiable.
• Protect the streak, not the volume. On bad days, do 10 minutes of reviews and zero new words. A 10-minute day costs almost nothing; a skipped week costs a painful backlog. The daily study page exists for exactly those days.
• Recall in both directions eventually. Recognition (Japanese → meaning) is enough for the exam, but adding a small production pass (meaning → Japanese) for Band-1 words roughly doubles their durability, because retrieval in either direction strengthens the same memory trace.

7. A sample week at N1 pace

• Mon–Fri (60–70 min/day): 30–40 min SRS reviews → 11 new Band-1 words from the N1 list → one NHK article with mining.
• Saturday (90 min): synonym-cluster session (build 2–3 clusters from the week's words) + longer-form reading.
• Sunday (60 min): weekly sweep of all new cards + leech surgery on anything failed twice this week.

That is roughly 8 hours/week on vocabulary alone — realistic for N1, where vocabulary quietly determines your reading speed and your listening comprehension ceiling at the same time. Coming straight from an N2 pass? Read the N2 study plan first to check your foundation, and if you're weighing N1 against a JPT 900 target for work in Korea, see JLPT vs JPT.