Hazard perception is 14 clips, 15 developing hazards and 75 points, with 44 needed to pass. Almost everyone who fails it fails for one of two reasons: they click at potential hazards instead of developing ones, or they click so often that the test scores them nothing. Both are fixable, and neither is about reaction speed.
Of the two halves of the theory test, this is the one people arrive at with the least idea of what is being measured. The multiple choice section is recognisably an exam. The clips look like a reaction test, so people prepare for them like a reaction test, and then fail them.
Hazard perception is not testing how fast you can click. It is testing whether you can tell the difference between something that might become a problem and something that has started to become one, and whether you notice the change at the moment it happens. That distinction is the entire test, and once it is clear the rest of the rules make sense.
The Format, Briefly
The mechanics are published and short.
| Element | Detail |
|---|---|
| Clips | 14 |
| Developing hazards | 15 |
| Maximum score per hazard | 5 |
| Points available | 75 |
| Pass mark | 44 |
Source: DVSA, the hazard perception test and pass mark and result. Correct at 1 September 2026.
Every clip contains at least one developing hazard, and one clip contains two, which is why 14 clips produce 15 hazards. You get one attempt at each clip and you cannot review or change your responses. There is no flagging and no going back, unlike the multiple choice section.
Potential Versus Developing: The Whole Test
A driving scene is full of things that could go wrong. A parked van. A cyclist ahead. A pedestrian on the kerb. A side road on the left. Every one of those is a potential hazard, and a driver should be aware of all of them at once.
A developing hazard is a potential hazard that has begun to change into something you must respond to. The word doing the work is "developing". Something in the scene alters, and that alteration is what you are clicking at.
- A parked van is potential. Its indicator coming on, or its front wheels turning out, is developing.
- A pedestrian standing on the kerb is potential. The same pedestrian turning towards the road and stepping off it is developing.
- A cyclist riding straight ahead is potential. The cyclist glancing over their shoulder, or drifting out around a drain, is developing.
- A side road is potential. A car nosing out of it is developing.
- A queue of stopped traffic ahead is potential. Its brake lights coming on, or a door opening, is developing.
In every pair, the second version requires you to do something: ease off, cover the brake, move out, or plan a different line. That requirement is the definition. If nothing about your driving would need to change, it has not developed yet.
Why Early Clicks Score More
The DVSA is explicit that to get a high score you should click as soon as you see the hazard starting to develop. Five points are available for each hazard, and the score you get reflects how early in that development you responded.
It is worth being clear about what the DVSA does not publish: the exact timings of the scoring bands. Plenty of websites will tell you a window opens at a precise second and closes at another. Those numbers are not in any official source, and treating them as fact leads people to try to time clicks rather than watch the road, which is the opposite of the skill being tested.
What you can rely on is the direction: earlier scores higher, and once the hazard is fully developed and obvious, the points have gone. This is why the test rewards drivers who read a scene early.
A driver who notices the van's indicator scores well. A driver who notices the van when it is already halfway out of the space has responded too late to score, even though they did respond.
That is also, incidentally, the reason the test exists in this form. Late observation is not a scoring quirk, it is the fault that ends more practical tests than any other, and it sits at the top of the DVSA's published reasons people fail the driving test.
Why Clicking Constantly Scores Zero
If you click continuously or in a pattern, you will not score anything on that clip.
This is not a glitch and it is not the software being harsh. It is a deliberate anti-cheat response, and the reasoning is simple. A candidate who clicks every half second will land a click inside every scoring window on every clip, and would score full marks while demonstrating nothing at all.
The test detects that behaviour and zeroes the clip, so the strategy that looks like it guarantees a pass actually guarantees a fail.
The consequence people miss is that this cuts both ways. You do not need to be metronomic to trigger it. A nervous candidate who clicks at everything that moves, several times per hazard, out of anxiety about missing one, can produce something the test reads as a pattern. Losing an entire clip to over-clicking costs 5 points, and you only have 31 points of slack across the whole test.
The practical question is how many clicks are safe, and the answer is that you can click as often as you could justify.
- One click at the moment something changes, which is what a driver actually does, is enough
- A second click shortly afterwards, if you want the reassurance, is ordinary and unremarkable
- Ten clicks spread across a clip in the hope that one of them lands is not, and it is exactly what the test is built to catch
- If you could explain out loud why each click happened, you will never trip the anti-cheat response
What Actually Goes Wrong
Nearly every hazard perception failure is one of four things, and none of them are slow reflexes.
- Clicking at potential hazards. The most common by far. The candidate is alert and observant, clicks at everything worth noticing, and scores on almost none of it because they clicked before anything changed.
- Clicking too late. The hazard is seen, but only once it is unmissable. This scores low or nothing and usually reflects looking at the wrong part of the frame rather than looking slowly.
- Over-clicking. Anxiety converts into volume, the pattern is detected, and clips are zeroed.
- Freezing on one clip. A candidate who thinks they missed a hazard spends the next clip thinking about the last one. Since there is one attempt per clip and no way back, this compounds quickly.
Note what is absent from that list. Not knowing the Highway Code well enough is not on it, and neither is reaction time in any physical sense. The clips reward a way of looking, and a way of looking is trained by repetition rather than by reading.
How To Practise Properly
With the official material
The official DVSA revision materials include hazard perception practice, and that is the sensible place to start because the clips are built to the same specification as the test. Practise in short sessions, several times a week, rather than an hour the night before. Recognition is a skill that consolidates over time.
The single most useful habit is to review, not just score. After each practice clip, say out loud what the developing hazard was and what changed. If you cannot name the change, you were guessing, and a score achieved by guessing tells you nothing about the real test.
On real roads
Clips are a simulation of something you can practise directly, and this is where lessons and theory revision stop being separate activities. Commentary driving, where you narrate what you can see and what you expect it to do, trains exactly the discrimination the test measures. It feels awkward for the first ten minutes and then becomes automatic.
East London traffic is unusually good material for this. A run along Barking Road or Mare Street presents parked vehicles with drivers in them, cyclists in and out of narrow lanes, buses pulling out from stops, and pedestrians stepping between stationary traffic, which is the entire vocabulary of the clips at a far higher density than the clips themselves. Learners here often find the clips feel slow after a few weeks of lessons.
If you are not driving yet, do it as a passenger. It works from the front seat of a bus.
The Points Arithmetic Worth Knowing
You need 44 of 75, which is a little under 59%. Averaging just under 3 points per hazard passes. That is a reassuring number in one direction and a warning in another.
The reassurance is that you do not need to be early on every hazard. Consistent mid-range scoring passes comfortably, so a single hazard you scored nothing on is not fatal.
The warning is that whole zeroed clips are what break the arithmetic. Miss a hazard entirely and you are down 5 points from a budget of 31. Trip the anti-cheat response on the clip with two hazards in it and you are down 10 in one go. Failures are usually built from a handful of zeros rather than from being slightly late across the board.
Where This Sits In The Rest Of The Test
Both halves have to be passed in the same sitting, so the clips cannot be treated as the optional part. The full structure of the test, including the multiple choice half and what the day looks like, is in how to pass the theory test first time, and the revision method that works for both halves is in theory test revision that actually works.
If you are still working out the order of the whole process, the complete guide to learning to drive in East London covers every stage. And if you want to train hazard recognition in the place it actually matters, that happens in automatic lessons rather than in front of a screen.
The fastest way to get better at spotting developing hazards is to spend time in traffic being asked what you can see, which is what a lesson in dense East London traffic already involves.
Start the beginner courseQuestions about the clips
What is the difference between a potential and a developing hazard?
A potential hazard is anything that could cause a problem, such as a parked van or a pedestrian on the pavement. A developing hazard is one that has actually started to, such as that van indicating or that pedestrian stepping into the road. Only developing hazards score.
Why did I score zero on a clip I definitely clicked on?
Almost always because you clicked too often. Clicking continuously or in a pattern scores nothing for that clip, by design, because otherwise rapid clicking would guarantee full marks. The other possibility is that every click landed before the hazard began to develop.
How many times should I click per hazard?
Once when it develops, and a second time shortly after if you want the confirmation. That is normal and safe. What causes problems is clicking repeatedly throughout a clip in the hope of catching the window.
Can I go back and redo a clip?
No. You get one attempt at each clip and you cannot review or change your responses. This is the main way the hazard perception section differs from the multiple choice section, where you can flag questions and change answers freely.
Does the hazard perception test have a time window I can learn?
The DVSA does not publish the scoring band timings, and any site quoting exact seconds is guessing. What is published is that the earlier you click as the hazard develops, the more of the 5 points you get.
Hazard perception and real driving train the same skill, and the clips get noticeably easier once you have spent a few hours reading real roads from the driver's seat.
Start the beginner courseMore in The theory test
Published 1 September 2026. Prices and DVSA figures are checked when this page is revised. If anything here has changed, ring us and we will confirm it.