Annual kilometres of travel on roads per hospitalisation, by mode, Melbourne, 2012-2013 (source data: VISTA, TAC)

A new study by researchers at Monash University found the total number of deaths in Victoria from road traffic crashes fell over the period from 2007 to 2015. But it’s not all good news. Serious road traffic injuries – which far outnumber fatalities – didn’t fall; they were steady for all modes except cycling.

Cycling is particularly worrying as it was the only mode to experience an increase in the incidence of hospitalisations for road traffic-related major trauma over the period; they increased by a staggering 8% per year (see More cyclists are ending up in hospital with serious injuries, so we need to act now).

So if cycling injuries are increasing, is the conclusion to be drawn from the study, “don’t cycle”? The authors respond:

No, not at all. The health and economic benefits of cycling are well established…And while cycling-related injury rates are on the rise, they made up only 11% of serious road traffic injuries.

But is “only 11% of serious road traffic injuries” really nothing to worry about? That depends in large part on the mode share of cycling but one of the limitations of the study, as the authors acknowledge, is they don’t have data on the exposure of cyclists.

Fortunately, the Victorian integrated study of travel and activity (VISTA) provides an estimate of the number of kilometres cycled annually in Melbourne by mode. The Victorian Transport Accident Commission’s (TAC) road trauma database provides statistics on hospitalisations by mode for Melbourne. It’s only one city but it should give us a reasonable idea of what’s going on.

I averaged the serious injury numbers for 2012 and 2013 as that fits with the VISTA period. They show an average of 330 cyclists were hospitalised due to road crashes per year, accounting for 9.0% of TAC claims involving hospitalisation. But cyclists only accounted for 1.2% of total annual kilometres of travel, indicating cycling casualties were greatly over-represented. (1)

As the exhibit shows, on average cyclists suffered an injury requiring hospitalisation at over nine times the rate of persons travelling in a vehicle. In fact, that significantly overstates the risk associated with a car, because the VISTA data doesn’t permit motorcycles – the most dangerous mode – to be separated from cars.

But the risk associated with cycling is still very small. Consider that a Melbourne worker who makes the average one-way commute of 6.5 km every day will come close to cycling 3,000 km in a year. That compares very well with one cyclist being hospitalised in Melbourne on average every 1,185,000 km of riding.

Or use time cycled as the measure of exposure. The same Melbourne worker commutes by bicycle for circa 250 hours per year. Again, that compares well with one cyclist being seriously injured on average every 107,754 hours in the saddle (that’s every 12 years!). (2)

So why are so many people nervous about cycling on roads? That’s a complex issue but I think a key insight is that our tolerance for risk is extraordinarily low and getting lower. Even though the probability of serious injury when driving in Melbourne is miniscule, the value of a human life is very high; we consequently insist on spending large amounts to equip cars with elaborate safety gear and to build roads that reduce the probability and severity of crashes.

It follows that if that’s what we expect of cars, we aren’t likely to cycle in large numbers while prospective riders feel it’s unsafe compared to other modes. The priority must be to provide infrastructure and a regulatory environment that significantly boosts subjective safety.

And why are serious cycling injuries increasing so fast? Unfortunately we don’t have reliable evidence to answer this question. It could be a composition effect e.g. an increasing proportion of riders are “lycra louts” and/or highly inexperienced. My hunch is the number of cyclists is increasing faster than the available surveys suggest (see Is cycling really declining in Australia?).


  1. There were seven cycling fatalities on Melbourne roads over the two years from 1 January 2012 to 31 December 2013.
  2. When exposure is measured by time, on average cyclists suffer an injury requiring hospitalisation at three and a half times the rate of persons travelling in a vehicle.