Did the Wales 20mph Limit Cut Crashes by a Third?

Quick answer: In April to June 2026, 299 collisions and 378 casualties were recorded on 20mph and 30mph roads combined in Wales, according to the Welsh Government. That is 37% fewer collisions and 35% fewer casualties than in the same quarter in 2023, the last quarter before the default 20mph limit began. Deaths tell a narrower story: 19 people were killed on 20mph roads in Wales in 2025, the same number as in 2024.

What the new figures show

The figures come from the Welsh Government bulletin "Police recorded road collisions: April to June 2026 (provisional)", published on 29 October 2026. It reports police-recorded collisions under STATS19. Comparisons are with the same quarter in earlier years, because collisions are seasonal.

On all roads in Wales in the second quarter of 2026:

• 638 police-reported collisions, resulting in 856 casualties
• Collisions down 19% and casualties down 20% compared with the same quarter in 2025
• Collisions 27% lower and casualties 25% lower than the same quarter in 2023

On 20mph and 30mph roads combined, the roads affected by the default limit change:

• 299 collisions and 378 casualties in April to June 2026
• Compared with the same quarter in 2025 (383 collisions, 474 casualties): collisions down 22%, casualties down 20%
• Compared with the same quarter in 2023 (477 collisions, 584 casualties): collisions down 37%, casualties down 35%

The Welsh Government news release described these as the lowest figures ever recorded in this way, excluding data captured during the Covid pandemic. The Deputy Minister for Transport, Mark Hooper, said: "These latest figures are extremely encouraging and provide more evidence that the 20mph default speed limit is helping to make our roads safer." He also said the government would continue to monitor the trends closely as part of its ongoing evaluation. That wording matters: the minister presents the figures as evidence, not as final proof. Readers can see other British data on the UK road accidents pages, read other data articles in all articles, and see the latest reported crashes on accidents today.

Why 20mph and 30mph roads are counted together

The combined figure keeps the comparison fair. Most roads that were 30mph became 20mph on 17 September 2023. At the change, 20mph roads grew from approximately 870 kilometres to about 13,000 kilometres, while 30mph roads shrank from approximately 13,100 kilometres to about 980 kilometres.

Counting 20mph roads alone before and after would compare a small set of roads with a much larger one. Counting them together keeps the same set comparable. Every headline percentage here therefore says "20mph and 30mph roads combined".

The bulletin also notes that quarterly collisions have declined steadily over the last decade, so the fall since 2023 sits on top of a longer downward trend.

What the figures do not show

The bulletin measures collisions and casualties. It does not show a fall in deaths on 20mph roads.

The Welsh Government For the Record blog of 30 September 2026 reports 19 fatalities on 20mph roads in 2025, the same number as in 2024. The same blog gives deaths on all roads in Wales: 91 people killed in 2025, compared with 84 in 2024 and 96 in 2022, the year before the default limit began. The blog was written to correct a claim that fatalities had increased since the limit began; that claim was misleading, because the figure quoted referred to all roads, not just 20mph roads. It does not show that overall road deaths fell.

Two caveats come from the bulletin. First, the 2026 figures are provisional and subject to revision. Second, the killed-or-seriously-injured total for the quarter, 294 casualties, uses adjusted severity data and is not comparable with earlier unadjusted publications. This article therefore does not trend the 294 figure.

The adjustment exists because forces moved at different times to injury-based reporting, in which severity is set from a predefined list of injuries. Dyfed-Powys moved in May 2023 and South Wales Police in November 2025; the other two Welsh forces have yet to move. South Wales adjusted data, based on only November 2025 to June 2026, is expected to be revised more than other figures. The reporting change does not affect total collisions, casualties or fatalities.

Cyclists, and the wider picture

Road.cc, reporting the same Welsh Government release, states that there were 45 crashes involving cyclists in Wales between April and June 2026, with 15 riders seriously injured and no deaths. In the same quarter in 2023 there were 78 collisions and 28 serious injuries. Those cyclist figures cover all roads in Wales, not 20mph and 30mph roads only, and the 37% and 35% falls above must not be attached to them. The blog also reports casualties on 20mph and 30mph roads combined are 23% lower since the limit began, a separate comparison from the quarterly figures above.

Taken together, the evidence supports a careful conclusion. Collisions and casualties are down sharply on the last quarter before the change. Deaths on 20mph roads have not fallen, and deaths on all Welsh roads rose in 2025. The death figures may be too small and variable to show a trend yet.

How these numbers were counted

All quarterly counts and percentages come from the Welsh Government bulletin "Police recorded road collisions: April to June 2026 (provisional)" (29 October 2026) and the Welsh Government news release of the same date for the 19% fall in collisions on all roads. The bulletin counts police-recorded collisions and casualties in Wales under STATS19. Figures for 2026 are provisional and may be revised. Comparisons are with the same quarter in 2025 and 2023, because collisions are seasonal; 2023 is the last second quarter before the default limit began on 17 September 2023. The combined grouping keeps the same set of roads comparable before and after the change. Death figures come from the Welsh Government For the Record blog of 30 September 2026. Cyclist figures are attributed to road.cc reporting of the release, and cover all roads in Wales. The killed-or-seriously-injured total uses adjusted severity data that is not comparable with earlier unadjusted publications.