Mock tests are the most misused tool in exam preparation. Taken the usual way — casually, untimed, glanced at, score noted — they measure your Japanese without improving it.
Used properly, they are the single most productive hour in your week.
The three jobs a mock test does
Be clear about which one you are doing, because they need different conditions.
- Diagnosis. Which section is costing you points?
- Rehearsal. Can you perform under sixty minutes of pressure on a computer?
- Go/no-go. Are you ready to book the real thing?
A relaxed untimed run is fine for diagnosis and useless for the other two.
Conditions, for anything except pure diagnosis
- Full length. Not half.
- Timed. One sixty-minute clock for the whole test, because the sections are not separately timed.
- On a computer. The real test is CBT, and the screen is a variable worth removing in advance.
- No pausing, no dictionary, no phone.
- Headphones for the listening items, once through.
Anything less rehearses a test you are not going to sit.
The part people skip
The review is the study. The test is just how you find out what to review.
An hour spent taking a mock and five minutes checking the score is an hour mostly wasted. Budget as long for the review as for the test.
For every question you got wrong, decide which of these it was:
| Cause | The actual fix |
|---|---|
| I did not know the word | Vocabulary. Add it, with its confusable neighbour. |
| I knew the words, misread the structure | Grammar. Usually a particle or a negative. |
| I ran out of time | Timing. Not a Japanese problem at all. |
| I knew it and clicked wrong | Carelessness. Slow down at the point of answering. |
These four have completely different remedies, and the most common mistake in exam preparation is treating all four as vocabulary problems. A timing failure does not respond to more words.
Also review the questions you got right but guessed. They are failures that happened to score. If you do not mark them at the time, you will never find them.
How often
- Early on: once, at the very start, before you have studied anything. It tells you where to aim, and everything after it depends on that answer.
- Middle weeks: roughly one every two weeks, to check the plan is working.
- Final fortnight: twice a week, under full conditions.
More than that and you are testing rather than learning. There is nothing to measure if nothing has changed since Tuesday.
What a mock score is not
It is not a predicted score. A short practice test cannot estimate a 250-point scaled exam, and anyone telling you that twelve or twenty-four questions predict your result is being careless with something expensive.
Use mocks for readiness and direction, not prediction. Our short free practice test deliberately reports readiness with a per-section breakdown rather than a projected number, for exactly that reason.
A full-length mock can go further, and ours does: the free exam demo is 48 questions under exam conditions and reports an estimated 10–250 score with the CEFR band it falls in. That estimate is a straight line drawn onto the official scale, not the Japan Foundation’s own equating, which is not published — so read it as direction with a number attached, and trust two mocks in the same week rather than either one alone.
The booking rule
When mocks have done their job, they will tell you clearly:
Two full-length, timed tests in the same week, both at 200 or above, on a computer. Then book.
One good score is noise — you were fresh, or the topics suited you. Two in the same week is a signal. Below that, a test date is a fee rather than a plan, and what a retake costs is worth reading before you spend it.