Mock tests are the most misused tool in JLPT preparation. Candidates either sit none until the week before, or sit one every weekend and wonder why their score stopped moving.
Neither is a schedule. Here is one.
The shape
Counting backwards from the test date:
| When | What | Why |
|---|---|---|
| 4-5 months out | One full mock | A baseline, and a shock worth having early |
| 3 months out | One full mock | Confirms the plan is working |
| 8 weeks out | One full mock | Last chance to change direction |
| 5 weeks out | Full mock, exam timing | Now under real conditions |
| 3 weeks out | Full mock, exam timing | Fine-tuning, not learning |
| 10 days out | Final full mock | Then stop |
Six mocks across five months. Between them, sit individual sections under time pressure - a reading section here, a listening section there - which cost 40 minutes rather than three hours.
Why not more often
A JLPT mock costs about three hours to sit and two to review properly. Five hours weekly is a large share of most candidates’ study time, and measurement does not teach.
A mock tells you what to study. It does not do the studying. If you sit one weekly you will have excellent data about a vocabulary base that is not growing.
Why not fewer
The first mock is always a shock, whatever the level. Better to have that shock in month one, when there is time to respond, than three weeks out.
It also catches structural problems early: pacing, the sectional minimums, and the fact that the listening section comes last when you are tired.
Sit them properly
Full length, timed, in one sitting, no dictionary, no pausing. A mock done in pieces across an evening measures nothing.
Match the real thing: the same time of day if you can, and the sections in the exam’s order. The listening section’s difficulty is partly a fatigue effect, and practising it fresh hides that.
Review is where the value is
Two hours, the same day or the next. Sort every question into knew it / guessed it / got it wrong / do not understand why. Guesses count as errors. Build the error list, review it weekly.
The full method is the same one we use for JFT-Basic: how to review a mock test.
Watch the sectional minimums, not just the total
This is specific to JLPT and it changes what a mock means. A total well above the pass mark with one section below its minimum is a fail, and candidates who only look at the total do not see it coming. Sectional minimums explained.
Where to find papers
The official JLPT site publishes sample questions and past workbooks; they are the only definitively representative material. Using past papers covers how to get value from a limited supply.
If your real target is a work visa rather than a certificate, JFT-Basic is the cheaper and more frequent exam - our free exam demo is a full-length mock of that test, sat under its own rules.