Which JLPT Level Should You Take? A Diagnostic Guide for First-Timers
Which JLPT level should you take? It is a self-selection question, not an eligibility one: the test lets you register directly for any of its five levels, with no prerequisite and no penalty for skipping.12 So the real decision is not "which can I qualify for" but "which should I sit." This guide routes that decision through a diagnostic flowchart driven by three inputs: hours studied, your current grammar, vocab, and kanji counts, and your target use case. It then refines the result with the first-timer rule.
Overview
The Japanese-Language Proficiency Test (JLPT) offers five levels, from N5 (most basic) to N1 (most advanced).1 You choose your own level when you register, so picking well matters more than passing some gate.
Three things decide the right level for you: how much you have studied, what you can already do (grammar, vocabulary, kanji), and why you need the certificate at all.
This article maps those three inputs to a recommended level, then applies the most-repeated piece of first-timer advice: test one level lower than you think. It links out to J-Compass grammar, vocabulary, and kanji homes for the counts rather than re-teaching them here.
How JLPT level selection actually works
You pick your own level (no prerequisites)
The JLPT has five levels, N1 (most advanced) through N5 (most basic). You register directly for any one of them, and passing a lower level is not a prerequisite for sitting a higher one.12
The five-level scheme dates to a 2010 redesign. The test was rebuilt from four levels (old 1-kyū through 4-kyū) into the current five (N1 to N5), with N3 added between the old 3-kyū and 2-kyū.3
You sit one level per test date. There is no mechanism to attempt two levels on the same sitting, so the choice is deliberate, not a hedge.2
No one assigns you a level and no exam screens you into one. The entire selection burden is on you before you pay the registration fee, which is exactly why a structured diagnostic is worth the few minutes it takes.
What the JLPT does and does not measure
The JLPT measures language knowledge (vocabulary, grammar), reading, and listening. It has no speaking or writing section; every section is receptive.2
Each level certifies a band of receptive comprehension, defined by official can-do descriptors. That is what makes "match the level to the use case" the correct way to choose.1
The official one-line summary for each level reads as follows.1
| Level | Official can-do summary |
|---|---|
| N1 | "The ability to understand Japanese used in a variety of circumstances." |
| N2 | "The ability to understand Japanese used in everyday situations, and in a variety of circumstances to a certain degree." |
| N3 | "The ability to understand Japanese used in everyday situations to a certain degree." |
| N4 | "The ability to understand basic Japanese." |
| N5 | "The ability to understand some basic Japanese." |
The official summary expands each level into a reading skill and a listening skill.1
- N1 reads writings with logical complexity and abstract content across topics, such as newspaper editorials and critiques, and follows the writer's intent; listening covers coherent conversations, news reports, and lectures spoken at natural speed in a broad range of settings.1
- N2 reads clearly written materials such as newspaper and magazine articles and commentaries on a variety of topics; listening covers conversations and news reports spoken at nearly natural speed in everyday and varied settings.1
- N3 reads materials with concrete contents concerning everyday topics and grasps summary information such as newspaper headlines; listening follows coherent conversations in everyday situations spoken at near-natural speed.1
- N4 reads passages on familiar daily topics written in basic vocabulary and kanji; listening covers daily-life conversations spoken slowly.1
- N5 reads typical expressions and sentences written in hiragana, katakana, and basic kanji; listening picks up necessary information from short, slow daily-life conversations.1
The three inputs that decide your level
The recommendation rests on three variables. Each gets its own subsection below, and the flowchart that follows combines them.
Input 1: Hours studied (with honest variance)
No JLPT body publishes a study-hour figure. The Japan Foundation and JEES do not certify any such number, so every hours estimate you see traces to the Japanese Language Education Center or to school and community compilations.34
A simplified set of round numbers circulates widely as a planning rule of thumb: roughly 150 hours for N5, 300 for N4, 600 for N3, 1,000 for N2, and 2,000 or more for N1. Treat these as a memory aid, not a measurement.
The sourced bands are wider and, importantly, split by whether you already read Chinese characters. The figures below are cumulative from zero, for the post-2010 test.3
| Level | With kanji background | Without kanji background |
|---|---|---|
| N5 | 250–450 hr | 325–600 hr |
| N4 | 400–700 hr | 575–1,000 hr |
| N3 | 700–1,100 hr | 950–1,700 hr |
| N2 | 1,150–1,800 hr | 1,600–2,800 hr |
| N1 | 1,700–2,600 hr | 3,000–4,800 hr |
A second school source gives broadly compatible single-point figures, again split by background. It estimates N5 around 350 hours with a kanji background versus 400 to 500 without; N4 about 550 versus 800 to 1,000; N3 about 900 versus 1,325; N2 about 1,500 versus 2,200; and N1 about 2,150 versus 3,900 to 4,500.4
The split between the two columns is the honest caveat that most hours charts hide. Learners already literate in Chinese characters save several hundred hours per level. The gap widens toward N1, where the kanji and vocabulary load is heaviest; zero-background learners trend to the high end of every band.34
A Chinese-literate learner and a zero-background learner at the same hour count are not at the same level. Read your own column, not the headline average, or you will misjudge where you stand by hundreds of hours.
The numbers also assume reasonably efficient, active study. Passive exposure inflates the real-world clock, and the source itself says the totals vary a lot by learner.4
Input 2: Your current grammar, vocab, and kanji counts
These per-level counts are unofficial. After the 2010 redesign, the JLPT bodies stopped publishing a test content specification, so the counts you see are reverse-engineered from past papers and prep materials rather than official targets.5
The kanji counts are the firmer of the two. Widely cited cumulative figures run roughly 100 at N5, 300 at N4, 600 at N3, 1,000 at N2, and 2,000 at N1. That is a near-match for the full jōyō kanji list.6
Vocabulary counts rest on weaker secondary consensus and should be read as bands. The commonly cited cumulative set runs roughly 800 at N5, 1,500 at N4, 3,700 at N3, 6,000 at N2, and 10,000 at N1. JLPT Vocabulary by Level breaks down the new-words-per-level deltas behind these cumulative figures.
| Level | Kanji (cumulative, approx.) | Vocabulary (cumulative, approx.) |
|---|---|---|
| N5 | ~100 | ~800 |
| N4 | ~300 | ~1,500 |
| N3 | ~600 | ~3,700 |
| N2 | ~1,000 | ~6,000 |
| N1 | ~2,000 | ~10,000 |
Grammar-point counts rise sharply between N4 and N2, which is where many learners stall. For the actual word, kanji, and grammar inventories, J-Compass maintains dedicated homes such as How Many Kanji Do You Need? and How Many Japanese Words Do You Need to Be Fluent?. This guide curates and points to them rather than reproducing the lists.
Because no official list has existed since 2010, two reputable prep sites can disagree by hundreds of items at the same level. Use the counts to locate yourself roughly, never to decide a pass by a single word.5
Input 3: Your target use case
The cleanest way to choose a level is to work backward from what the certificate must unlock. The official merit page states several concrete benchmarks.7
For Japan's points-based system for highly skilled foreign professionals, passing N1 earns 15 points and N2 earns 10 points toward the 70-point threshold for preferential immigration treatment.7
Residence-status language benchmarks appear on the same official page:7
| Use case | Level that satisfies it |
|---|---|
| Student visa at a Japanese-language institution | N5 or higher |
| Specified Skilled Worker (i) | N4 or higher |
| Engineer / Specialist in Humanities / International Services | N2 or higher |
| Business Manager | N2 or higher |
| Highly Skilled Professional points | N2 = 10 pts, N1 = 15 pts |
For higher education, N2 or higher is the benchmark the official merit page cites for admission to Japanese higher-education institutions.7
For undergraduate admission specifically, Japanese universities primarily use the Examination for Japanese University Admission (EJU). It evaluates the Japanese-language and academic readiness of prospective international students.8 The EJU was adopted in 2002 as a replacement for the older JLPT-based admission route.9
Some licensing pathways set their own minimums. N1 is required for several overseas-licensed medical and nursing national-exam routes. EPA nurse and caregiver candidates need N5 or higher (Filipino), N4 or higher (Indonesian), or N3 or higher (Vietnamese).7
For a CV or office work, N2 is the practical résumé line implied by the residence-status benchmarks above. N3 is widely treated as an intermediate milestone but certifies fewer concrete privileges.7
The diagnostic flowchart
The chart below combines the three inputs into a single path. Start at the top with your strongest signal. If a use case forces a specific level (a visa or admission requirement), let it dominate, since a certificate you cannot use is wasted effort. Otherwise, route on hours and counts.
How to read the flowchart
The chart gives a starting recommendation, not a verdict. It points you at a candidate level; the next two sections refine it.
The first-timer rule may pull a borderline case down one level, and the mock-test check converts the estimate into evidence before you register.
The first-timer rule: test one level lower than you think
The most-repeated piece of first-timer advice is to register for the level below the one you believe you are ready for. This is community guidance, not an official rule, but it has a mechanical reason behind it.
Passing the JLPT requires meeting both the overall pass mark and a per-section minimum. One weak section can sink an otherwise-passing total.10 A learner with a strong reading score and a weak listening score can clear the overall mark and still fail.
The passing mechanics, out of 180 total points:10
| Level | Overall pass mark | Sectional minimum |
|---|---|---|
| N1 | 100 / 180 | 19 / 60 in each of the 3 sections |
| N2 | 90 / 180 | 19 / 60 in each of the 3 sections |
| N3 | 95 / 180 | 19 / 60 in each of the 3 sections |
| N4 | 90 / 180 | Lang. Knowledge + Reading 38/120; Listening 19/60 |
| N5 | 80 / 180 | Lang. Knowledge + Reading 38/120; Listening 19/60 |
To pass, the total must be at or above the overall pass mark and the score in each scoring section must be at or above that section's sectional minimum.10
Three other honest reasons favor aiming one level lower on a first attempt. Test-day time pressure on an unfamiliar format costs more than it does in practice. A clean pass carries real momentum and confidence value, and the registration fee is non-trivial per sitting. Read this as advice for first-timers and borderline cases, not a blanket "never aim high."
When NOT to drop a level
Dropping a level is a default, not a law. Three situations argue for sitting your target level anyway.
Sit the target level if you comfortably clear timed past papers at that level. The mock test has already supplied the evidence the rule is trying to protect.
Sit it if you are on a fixed visa or admission deadline and a lower certificate will not arrive in time to matter. Also sit it if the lower level certifies nothing you actually need. The use-case map in Input 3 is the test for that last one.7
Confirm with a mock test
Convert the estimate into evidence before you pay. Sit a timed official past paper or practice test for your candidate level and score it against the overall pass mark and every sectional minimum.10
A mock test settles the borderline cases that the flowchart and the first-timer rule only estimate. If you clear the marks with room to spare, hold your level. If a section dips below its minimum, that is the signal to drop one.
This guide defers the full scoring and section-structure detail to The JLPT Explained; the numbers above come from the official pass/fail and section-composition pages.102
Good to know
The JLPT has no speaking or writing section
The test is entirely receptive, covering language knowledge, reading, and listening, with no production component.2 A certificate therefore is not direct proof of conversational ability. J-Compass covers where the JLPT falls short in full. Choose your level for what the certificate has to do, not as a self-image of fluency.
Skipping levels is allowed but rarely free
Registration imposes no prerequisite, so a candidate can jump from N5 to N2 if they wish.12 The catch is that the level scheme is cumulative in content even though registration is not gated. Foundational grammar gaps surface on the way up. The choice is about evidence of readiness, not permission.
"Hours to level" numbers are planning aids, not promises
The hour bands are unofficial Japanese Language Education Center and school estimates.34 They assume efficient study and vary widely by learner and prior kanji knowledge. No JLPT body certifies them. Use them to plan a timeline, not to guarantee a result.
One level per test date
A candidate sits a single level per exam date and cannot hedge by taking two levels the same day.2 You have to make the decision deliberately at registration, which is the whole reason to run the diagnostic first.
See also
- The JLPT: Test Format, Scoring, and Registration
- How Long to Prepare for Each JLPT Level: Hours, Months, and Honest Caveats
- Alternatives to the JLPT: BJT, J.TEST, JFT-Basic, and Kanji Kentei
- How to Learn Japanese: The Complete Roadmap from Zero to Fluency
- JLPT Vocabulary by Level: How Many Words for N5 to N1
- How Many Kanji Do You Need? A Realistic Count