The official question bank holds 3518 questions (version 07.2026), and your exam draws 32 of them. Studying for theory is therefore not about seeing everything once. It is about reaching the point where you can justify the answer before you click it. Here is a plan that gets you there, and the reason it beats one long push the night before.
Why clicking through the whole bank does not work
Clicking through the bank builds recognition memory, while the exam tests working knowledge. Recognition sounds like “I have seen this question”; knowing the rule sounds like “I know why this answer is the correct one” - and only the second one helps with a situation that looks familiar but is not.
At a bank of that size, recognition is misleading for two reasons. First, the feeling of “I know this one” arrives faster than the answer does, and it is enough to make you stop thinking. You close the question in two seconds, convinced you know it, when all you retrieved was the sense of having met it before. Second, many questions differ by a single element: a different sign, a different arrow, the other vehicle coming from the other direction. Recognition then points you to the neighbouring question rather than the one on your screen - the answer you remember is correct, just not here.
Then there is the arithmetic. You need 68 points out of 74 to pass, and questions carry different weights, so the margin is narrow and does not absorb guesswork on the harder items. Study built on things looking familiar falls apart exactly where it costs the most.
What the three-phase plan looks like
The plan has three phases and each does a different job: the first organises the material by topic, the second removes your own mistakes, the third checks whether all of it holds up under time pressure. The order matters - phase three done too early teaches you nothing except that you are not ready yet.
| Phase | What you do | How you know you can move on |
|---|---|---|
| 1. Topic sweep | Work through the material topic by topic, not at random. Short batches from a single topic, and you read the explanation even when you answered correctly | Within a topic you can justify the answer before you see the options |
| 2. Mistake work | Repeat only what you got wrong. For every mistake you write down the rule, not the letter of the answer | The same mistake stops coming back across two consecutive passes through your list |
| 3. Full runs | The whole set under exam conditions: no pausing, no going back, no peeking at explanations mid-run | Several runs with a margin above the pass mark, not one good score |
Phase one - by topic, not at random
A random set mixes topics, so you switch context after every question and nothing settles into a whole. Go topic by topic: right of way on its own, signs on their own, speed and distance on their own. Inside a single topic you start to see that a dozen questions rest on the same rule, and that is the moment the rule sticks.
Do not race the clock at this stage. Read the explanation even when you were right - a correct answer without a reason is the same gap in knowledge as a mistake, only better dressed.
Phase two - repeat what you got wrong
The second phase is shorter and less pleasant, because you work only on material that has already beaten you once. That is exactly why it pays best: repeating questions you already know costs the same time and changes nothing in your result.
In practice you come back to your mistake list after a few days, not straight after making the mistakes. A question you redo a minute later costs nothing, because the answer is still sitting in short-term memory.
Phase three - full runs under exam conditions
A full run means 32 questions in 25 minutes, with no going back to questions you have already answered. This phase is not there to teach you the material; it is there to make the format ordinary: the questions come one after another in an order you do not control, nothing can be corrected, and the clock keeps running.
If your run score is clearly lower than your topic-practice score, the problem is the format rather than your knowledge - and more runs fix it, not another sweep through the bank.
Why several short sessions beat one long push
Several short sessions a week hold better than one long one, even when the total time is identical. The reason is simple: memory strengthens what it had to retrieve with effort, not what it has just finished reading.
When you repeat material immediately after reading it, the answer is still within reach and recalling it costs nothing. After a break, the same material comes back reluctantly, and that resistance is what leaves a trace. A session that opens with a few questions from last time is therefore worth more than the same quarter of an hour added to new material.
Cramming the day before has one further flaw that has nothing to do with memory: it leaves no room for phase two. You build your mistake list on the one day when there is no time left to do anything with it.
How to work from your mistakes
Your own list of mistakes is worth more than a percentage score, because the score tells you where you were and the list tells you what to do next. There is one method: for every mistake, write down the rule rather than the answer.
The difference is decisive. “In this question the answer is no” works exactly once, in that one question. “At an intersection of equivalent roads a rail vehicle has priority regardless of which direction it comes from” works in every tram question about such an intersection, including the ones you have not seen yet. The first goes into memory as a single point; the second goes in as something you can use.
That gives you the simplest readiness test for a single question: try to state the rule in one sentence. If you cannot, you do not know the question yet - however many times you have clicked the right answer in it.
Which topics pay back the most
Four areas pay back most: right of way, signs and signals, safe speed and distance, and behaviour towards pedestrians and cyclists. They recur throughout the bank because they describe situations that happen constantly in traffic, rather than definitions you can memorise once.
Weighting is on their side too. Questions are not scored equally, and those describing a real decision at an intersection or conduct towards a pedestrian weigh more than dictionary items. With a pass mark of 68 out of 74, an hour spent on right of way moves your result more than an hour spent on definitions - even though both look identical in your count of questions solved.
This does not mean the rest can be skipped. It means that if your week holds three sessions instead of five, these four topics are the ones that should survive.
How you know you are ready
There is one criterion: several full runs with a margin above the pass mark, done under exam conditions. Not one good score but a repeatable one - a single run above the threshold can be luck, while a repeatable one cannot, because every set of questions is different.
“Under exam conditions” has a concrete meaning here: the full number of questions, the full time limit, no pausing, no going back, no looking at explanations before the end. A run with explanations open measures something other than the exam and produces a score you cannot carry into the room.
The margin matters in practice. Sets differ in difficulty, and a score sitting exactly on the threshold means the draw decides whether you pass, not your preparation.
How long the whole thing takes
There is no single number of days or hours, because the time depends on two things nobody can estimate for you: how much you already know from everyday life, and how many sessions per week you can actually complete. Someone who has cycled through a Polish city for years and picked up the signs by observation starts from a different place than someone reading about priority at an equivalent intersection for the first time.
The theory course at a driving school organises the material and sets the order of topics, but the number of classes attended is not a measure of readiness - no more than the number of questions clicked. The measure is a repeatable score in full runs. So plan your study around the number of sessions you can hold for several weeks in a row, not around the date you would like to be finished.
A plan that works looks boring: a few short sessions a week, your own list of mistakes, and several full runs at the end. It looks boring because it spends its time on what you do not know yet instead of reminding you of what you already do.
