Word against word: how evidence makes or breaks a non-fault claim

However obvious fault feels at the roadside, it has to be proved afterwards. That is harder than most drivers expect.

The other driver may apologise at the scene and tell a different story a week later. Where the circumstances are not clear and there is no key evidence to support your version, their insurer can argue that both drivers were to blame, or that the accident was entirely your fault. Most disputed claims come down to the same thing: one driver's word against another's.

That is why we cannot take on every claim. We recover our costs from the at-fault driver's insurer, which is only possible when we can prove their driver was at fault. Sometimes that means a driver who genuinely was not at fault still cannot have their claim taken forward, not because we doubt them, but because there is nothing to prove it with.

This report uses six years of our own records to show how much difference evidence makes, at two points. The first is our decision on whether we can take the claim on. With key evidence we could take forward 71.2% of enquiries. With none, 49.0%, fewer than half.

The second is the at-fault driver's insurer's decision on whether to formally accept that their driver was at fault. On the claims we took forward, the insurer formally accepted liability on 68.7% of claims with key evidence and 52.5% of claims without.

Key evidence is rare, though. More than four in five enquiries arrived with no witness and no footage, and fewer than one in ten had a witness. Drivers must stop and exchange details by law, but nothing tells them to gather proof, and most do not find out how much it matters until it is too late.

Source: Auto Claims Assist proprietary claims data, 2020-2025

Source: Auto Claims Assist proprietary claims data, 2020-2025

A driver with key evidence was 45% more likely to have their claim taken forward: 71.2% against 49.0%.

Drivers who arrive with no key evidence still say the accident was not their fault. But with nothing to back that up, whether we can take the claim on depends on the circumstances alone. We can take it on when the circumstances make fault clear, such as a rear-end collision. That was the case for just under half of these enquiries (49.0%).

This is not about whether we believe the driver. When there is no key evidence and the circumstances are not clear, it is one driver's word against the other's, and the at-fault driver's insurer is likely to dispute liability. That is the main reason we cannot take a claim forward: if we expect the insurer to dispute liability and we have nothing to prove otherwise, we cannot recover our costs. The same applies to some claims with key evidence, where the evidence is not strong enough on its own to prove who was at fault.

Under the Road Traffic Act 1988, drivers must stop and exchange details after a collision. The law says nothing about gathering proof. That is left to the driver, in the few minutes after the impact, and it is what decides how strong the claim will be.

Three things stand out.

First, key evidence still matters after we take a claim on. The insurer accepted liability on 68.7% of claims with key evidence and 52.5% without, a gap of 16.2 percentage points. A driver with key evidence was 31% more likely to have liability accepted.

Second, the biggest difference is in the claims where no stance was given at all. Here the insurer dealt with the claim without ever saying whether their driver was at fault. That includes 477 claims, one in every 27 we took forward, settled on a "WOP" basis, meaning the insurer agrees to pay but does not admit their driver was at fault, so they can revisit the position if new evidence turns up. It rarely does. What the insurer is really saying is that they did not have enough to dispute the claim, but were not prepared to admit liability either.

That matters to the driver. A formal acceptance of liability puts on record that you were not at fault. You can give it to your own insurer to bring a renewal quote down, or to close a claim still sitting open on your policy. A claim settled without it leaves the question of fault open. Without key evidence, a claim was 78% more likely to be left that way: 44.9% against 25.2%. Whether an insurer accepts liability is ultimately their decision, but it is why we push for a formal acceptance rather than settling for the claim being paid.

Third, key evidence lets us take on claims we otherwise could not. Without key evidence, a claim only goes forward when the circumstances are clear-cut and there is almost nothing to argue about, such as a rear-end collision. With key evidence, we can take on riskier claims, where the circumstances are more complicated and fault would otherwise be open to argument, because the evidence can prove them. Those are naturally the claims an insurer is more likely to dispute. That is why disputes were slightly more common on claims with key evidence, 6.1% against 2.7%, even though there were fewer of them in total: 192 against 255. The dispute rate is not a weakness of the evidence. It reflects the riskier claims the evidence allowed us to take on.

Whatever position the insurer takes, one thing does not change for the driver. If we take a claim on and you keep to the terms of your agreement with us, what the insurer decides is our risk.

Source: Auto Claims Assist proprietary claims data, 2020-2025

A witness came out ahead of video footage, both on the enquiries we could take forward and on the claims where the insurer accepted liability. Where the driver had a witness but no video footage, three in four enquiries were taken forward (75.7%). Where they had dashcam or CCTV footage but no witness, two in three (66.2%). Once the claim was taken on, the insurer accepted liability on 73.3% of the claims backed by a witness and 62.9% of those backed by video footage.

Video footage still made a big difference on its own. Compared with an enquiry with no key evidence, an enquiry with footage but no witness was 35% more likely to be taken forward and, once taken on, 20% more likely to have liability accepted, and 26% less likely to be left with no stance at all.

A witness came out slightly ahead in the numbers, but the two are not the same kind of evidence. Whether you have a witness depends on whether somebody happened to see the accident and stopped, which is out of your hands. A dashcam is in your hands. You can get ahead and plan for the worst by fitting one before anything happens: it records every journey, so if you are ever hit, the footage is already there.

Source: Auto Claims Assist proprietary claims data, 2020-2025

More than four in five drivers had no key evidence to back up their account. Fewer than one in ten had a witness.

Many drivers are left without key evidence after an accident because they never expected to need it. Fitting a dashcam may not have crossed their mind, while shock at the scene can make it easy to overlook speaking to witnesses, taking their details or checking for nearby CCTV that might have captured what happened.

Source: Auto Claims Assist proprietary claims data, 2020-2025

Dashcam footage came out ahead of CCTV on the enquiries we could take forward, 70.3% against 66.3%, and the likely reason is what each one captures. A dashcam records from the driver's own car, so it shows the lead-up to the accident as well as the accident itself: where each car was in the road, and what each driver did. CCTV may only catch part of the accident, or from an angle where the road markings and positions are not clear.

On whether the insurer then accepted liability, the two were almost identical: 63.6% of the claims with dashcam footage and 64.5% of those with CCTV. Once a claim was taken on, the two performed the same.

Afterwards: get independent support before accepting anyone's version of events. A claim with little evidence can still go forward, but it is hard for a driver to do alone against an insurer's claims department. And the work does not stop once a claim is taken on: we still have to get the insurer to accept liability.