It has been an awful year for cycling safety in Australia.
Last week, a 32-year-old cyclist died after a car collided with a group of five cyclists in Canberra, while another cyclist was critically injured.
It is unfortunate that it often takes a tragedy like this to draw attention to broader safety trends. And speaking of safety trends, the latest figures for Australian cyclists are particularly concerning.
In the 12 months to July 2026, 57 cyclists died on Australian roads. That is 54 per higher than the 37 deaths recorded over the previous 12 months.
The longer trend makes the latest figure even more striking. In the equivalent 12-month period ending July 2017, only 26 cyclists died.
Since then, the annual count has fluctuated between 33 and 46 deaths. It has now jumped to 57 and that is more than double the 2017 figure and 24 per cent above the previous decade-high of 46.
Could greater exposure be the main reason? More people cycling, or cycling more often, would naturally create more likelihoods for serious crashes.
The latest national participation data does not support this hypothesis as it doesn't show a particularly distinct cycling boom.
In 2025, about 15 per cent of Australians, or 4.11 million people, cycled in a typical week. That was virtually unchanged from 2023 and below the 18.2 per cent recorded in 2021.
The 2025 survey describes weekly cycling participation as steady, with riding frequency also stable at around three days per week.
Cyclists are not the only road users for whom the longer-term picture is concerning.
Over the same period, deaths among drivers have remained broadly flat, from 564 in the 12 months to July 2017 to 584 in 2026. Passenger deaths declined from 222 to 202.
The picture for vulnerable road users is different. Pedestrian deaths increased from 169 to 189 over the same period, while cyclist deaths more than doubled from 26 to 57.
It appears that improvements in road safety have not benefited all road users equally. In fact, it appears that the rising road trauma in the country is driven primarily by the rise in vulnerable road user deaths.
As for the tragic Canberra crash, we don't know all the details and exact circumstances yet, but one aspect is particularly striking and that's the fact that the cyclist who died was not riding alone. She was part of a group of five.
There is evidence of a "safety in numbers" effect in cycling, meaning the risk to an individual cyclist tends to decline as the number of cyclists increases.
Australian research has observed this effect, including at a Melbourne intersection where potential conflicts with vehicles decreased as cyclist numbers grew.
There's a lot of discussions around cycling infrastructure. But we should note that improving cyclist safety is not exclusively about improving the infrastructure. The attitudes and behaviour of drivers who share the road with cyclists also matter just as much.
An Australian study of nearly 2000 drivers found those who also cycled themselves were 1.5 times more likely to report safe driving behaviours around cyclists. Non-cycling drivers also expressed more negative attitudes towards cyclists.
Another Australian study found that dehumanising views of cyclists were associated with self-reported aggression towards them.
And among 3769 Queensland drivers, about one-third reported not consistently complying with minimum passing-distance rules. Most notably non-compliance was related more strongly to attitudes than demographic characteristics.
More recently, a survey of more than 2000 Australian and New Zealand cyclists found 87 per cent perceived stigma towards cyclists and 99 per cent reported having experienced conflict with drivers.
Cyclist safety requires attention to and investment in both the physical as well as the social infrastructure.
With respect to the alarming fatality statistics that I mentioned earlier, there's a limit to what these statistics can tell us. We can see that cyclist deaths have risen but we cannot readily determine why.
Finding these critical answers that can inform policy requires granular crash-level data, for both fatal and non-fatal crashes, that includes key variables determining the circumstances in which they occurred.
Victoria is one of the only states that makes detailed crash data publicly available.
In our recent analysis of pedestrian fatalities in Victoria, for example, this level of data availability allowed us to identify increasing vehicle size as an important contributor to the rise in pedestrian deaths. We estimated that had average vehicle size remained unchanged, pedestrian fatalities would have been nearly 9 per cent lower over the past decade.
There is a genuine need for similar access to detailed crash data nationally, while protecting privacy through de-identification is a default assumption. Without such data, every now and then we detect troubling trends such as this one for cyclists, but our ability to understand what is driving them remains unnecessarily limited.
We must do better than that in our diagnoses. That would be an important step towards reducing road trauma, especially for vulnerable road users, who seem to have become increasingly vulnerable on Australian roads.
