What Score Do You Need to Pass the JLPT?

To pass the JLPT, you need a total at or above the pass mark for your level (80/180 at N5, 90 at N4, 95 at N3, 90 at N2, 100 at N1) and at least 19/60 in every scoring section (38/120 for the combined N4/N5 section). That second condition is where the surprises live: we enumerated every possible score profile, and at the N2 pass mark, 78.7% of the profiles that reach 90 still fail because one section falls short.

This article gives you the official numbers in one table, then shows what they mean in practice. For each level’s format and content, see the individual guides, starting with the JLPT N5 guide.

What score do you need to pass each JLPT level?

Every level is scored out of 180, but the sections differ. N1, N2 and N3 have three scoring sections of 60 points each: Language Knowledge (vocabulary and grammar), Reading and Listening. N4 and N5 have two: Language Knowledge and Reading combined (0 to 120) and Listening (0 to 60).

Source: the Japan Foundation’s official scoring and pass/fail page, in force since the July 2010 test (December 2010 for N4 and N5).

Two details are worth noticing. First, N3 asks for a higher total than N2 (95 against 90), which looks backwards until you learn that each level has its own scale. Second, both minimums sit at the same proportion: 19 is 31.7% of 60, and 38 is 31.7% of 120.

Why isn’t the total score enough on its own?

Because the JLPT applies two tests at once, and failing either one is a fail. The official rule is blunt:

“If there is even one scoring section where the score is below the sectional pass mark, examinees are determined to have failed, no matter how high the total score he/she might have.”

In other words, the total checks that you know enough Japanese overall, and the section minimums check that you have no blind spot. A candidate who reads brilliantly but cannot follow the audio fails, even with a total that would pass most levels comfortably.

Missing a section entirely is also a fail. The score report then shows no scores at all, not even for the sections you did take.

How often does a passing total still fail?

We wrote a script, tools/jlpt_score_space.py, that lists every score profile the rules allow and checks each one against the official pass conditions. JLPT scores are whole numbers, so the space is finite: 226,981 profiles at N1, N2 and N3 (61 possible scores in each of three sections), and 7,381 at N4 and N5 (121 × 61). Before counting anything, the script tests the pass/fail rule on ten hand-made cases and re-reads the live jlpt.jp page to confirm the pass marks have not changed.

Here is what it found among the profiles whose total reaches the pass mark:

To make that concrete: an N2 report of 52, 20 and 18 adds up to exactly 90, the pass mark, and fails, because 18 is one point under the Listening minimum. A report of 30, 30 and 30 has the same total and passes easily. The total is identical; only the shape differs.

The pattern is consistent: the closer your total is to the pass mark, the more of the possible profiles fail on a section. At N2, the share of failing profiles drops from 78.7% at exactly 90 to 63.5% at 100, 47.7% at 110, 30.1% at 120 and 10.2% at 130.

N4 and N5 show the lowest shares because they have only two sections, and the combined one is worth 120 points. With fewer places for a weak score to hide, fewer of the profiles that reach the total are sunk by a single section. N2 shows the highest share because its pass mark is the lowest of the three-section levels, so more low-scoring shapes still reach it.

There is also a hard ceiling. The highest failing total at every level is 138, for example 60, 60 and 18. Any total of 139 or more always passes, because you cannot lose a section and still collect that many points elsewhere.

Why are these profiles and not pass rates?

Because we counted possible score reports, not people. Every profile counts once, including extremes like 60 in Reading and 0 in Listening that almost nobody produces. Real candidates cluster: someone strong enough to score 55 in Reading usually does not score 10 in Listening.

The JLPT also uses scaled scores, not raw points. In the words of the official FAQ:

“The JLPT uses a scaled score system based on “Item Response Theory” and does not total allocated points of correctly answered questions.”

The Japan Foundation’s explanation of scaled scores adds that the score depends on your answering pattern, which questions you got right, so two people with the same number of correct answers can receive different scores. Our enumeration describes the scoring rule, not the distribution of real results. What it proves is structural: a total near the pass mark is compatible with a large number of failing reports, so the total alone cannot reassure you.

What score should you actually aim for?

Aim for balance first and total second. If you split each pass mark evenly across the sections, you get a useful baseline:

A balanced candidate sitting exactly on the pass mark already clears every minimum by a comfortable margin. The risk comes from lopsided profiles: the most lopsided way to pass N2 with exactly 90 is 19, 19 and 52, one point above the floor twice.

In practice, three habits protect you:

1. Find your weakest section early. Take a timed practice test per section and look at the lowest score, not the average. If most of your study is textbooks and flashcards, check Listening first: text-based study barely trains it.
2. Train the weak section with volume, not tricks. Listening and reading speed respond to daily exposure far more than to test strategies. This is the same reason the N4 to N3 jump punishes flashcard-only study.
3. Target a buffer above the pass mark. Our table shows failing profiles thinning out quickly as totals rise. A total in the 120s leaves far less room for a hidden weak section than a total of 91.

If you are preparing a specific level, the JLPT N3 guide and the JLPT N2 guide break down each section’s question types, so you know what the minimum is actually measuring.

Where does Shinobi fit?

The section minimums punish a single blind spot, and for most learners that blind spot is comprehension speed: reading long passages in time, and following audio without a pause button. Shinobi is built for exactly that half of the exam. Its 455 graded stories run from pre-N5 to N2, each with native audio for listening practice, furigana and tap-to-translate for reading without breaking flow, and quizzes after every story.

Start with the stories at your target level, for example the N3 stories, and treat them as daily training for both the Reading and Listening floors.

What’s the bottom line?

The JLPT pass mark is only half of the rule. You also need 19/60 in every section (38/120 for the N4/N5 combined section), and a total close to the pass mark is compatible with many failing profiles: 78.7% of possible reports at exactly 90 on N2 fail a section. Only 139 or more is safe by arithmetic alone. Build a balanced score, find your weakest section early, and aim well above the line rather than onto it.