Home page

Main topics
News Headlines

Frequently asked Questions
For Policy Makers

Research evidence
Misleading claims
Helmet laws

Search Engine

New Zealand
Other countries

Full index

Policy statement
Register as a supporter

Download this page

An introduction to the objective risk assessment of cycling

A look at risk when cycling in Great Britain, France, the Netherlands and Denmark

It is often thought that cycling must be dangerous because cyclists are unprotected road users sharing space with motor traffic. This view has long been justified by data showing that, for example, in Great Britain the fatality rate per 100 million kilometres is, on average, 12 times greater for cyclists than for motorists .

In other countries, however, similar comparisons have found a much closer level of risk between cycling and motoring. For instance, a study in France by INRETS (the French national research organisation) found that the risk per kilometre travelled was about the same for cyclists and drivers (Carre, 1995). This reflected a better standard of safety for French cyclists, and a much poorer standard of safety for French drivers, relative to Britain.

Likewise in the Netherlands, risk comparison has shown similar or lower risks per kilometre travelled for cyclists younger than 50 years old relative to motorists (Carre, 1992).

During the 1980s, the UK Transport and Road Research Laboratory (TRRL) carried out a comprehensive study of risk for all road users in Great Britain (Morgan 1988). It noted that relative risk between different classes of road user depended on the basis for comparison. One could consider the risk per kilometre travelled, the risk per hour of travel, the risk per average trip, or the risk per year. These various methods yielded greatly differing results.

Comparison by risk per kilometre strongly favoured the traditional view that the motor car was much safer than either cycling or walking. Pedestrians had the highest fatality rate per 100 million kilometres, followed by cyclists, with drivers having a rate almost twenty times less than for pedestrians.

In contrast, the risk of fatality per year showed very little difference amongst the road users, although pedestrian risk was lower than for cyclists and drivers. This disparity may be explained by the fact that drivers are much more mobile than pedestrians and cyclists. A driver might well enjoy one twelfth the fatality rate per mile of a cyclist, but a cyclist only covers one twelfth the distance per year, so the accumulated risk per year is about the same. The TRRL research would have provided valuable guidance for an understanding of cyclist safety, but it was never published. Its conclusions remained unheard, and the safety debate uninformed..

If risk assessment is to be meaningful, the most appropriate basis for comparison must be chosen. The risk per hour of use is recommended. The justification is as follows. Long term monitoring of personal travel by the UK National Transport Survey and other international evidence (Shafer, 2000) shows that the time spent in daily travel is fairly constant over time, at about 1.2 hours per capita per day. The main reason why national mileage travelled has increased over time, with consequences for traffic congestion, is that car ownership has steadily increased. It is not because the overall time spent travelling by the population has increased since the 1970s.

If we are to consider people changing from motorised travel to walking and cycling (to benefit public health, reduce road danger and oil dependency) then we should not expect much change in the time spent travelling. Rather, the overall mobility of the population would reduce. The implications for road safety would then be best evaluated by considering risk per hour of travel in walking and cycling – and, further, the risk per hour of imposing death or serious injury on third parties.

When assessing risk, it is also important to recognise that pedestrians, cyclists and motorists are different sub-populations within an overall national population. For instance, in Britain, the average age of serious cyclist casualties is 23-25 years, whilst the average age of serious motorist casualties is 35 years. In addition, cycling in Britain is dominated by teenage boys and young men, the most injury-prone groups of all road users. Thus the aggregate risk calculated for cyclists will be heavily loaded by these risk factors. A higher average risk for cyclists viz-a-viz motorists does not mean that cycling is really more dangerous than driving when a like-for-like comparison is made.

Two studies have sought to pull together international data to make a comparison of risk in walking, cycling and motoring (Wardlaw, 2002; Krag, 2005). Both of these studies have found that pedestrians face higher risks per distance travelled than cyclists.

Wardlaw, 2002 noted that on a risk per hour basis, the comparison of risk between cyclists and motorists gave greatly varying results depending on the country in question. For instance, cycling is safer per hour than motoring in France, but the reverse is true in Britain. This is interesting because French cyclists share road space with motor traffic just as in Britain. The very different outcome could be related to the greater popularity and social profile of cycling in France, the greater involvement of experienced adults in cycling, and legal presumption – in France, a motorist is presumed culpable in a collision with a cyclist.

Krag, 2005 analysed the severity of injury of road users admitted to hospital in Odense, Denmark following road accidents. The analysis showed that cyclists were generally less gravely injured than other road users and were detained for less time in hospital. For instance, the average periods of detention in hospital were:

Pedestrian 3.2 days
Cyclist 0.5 days
Motor vehicle occupant 1.0 days

The percentage of more serious admissions, with AIS scores of 3 or more, were:

Pedestrian 9%
Cyclist 3%
Motor vehicle occupant 4%

(Note that Denmark is a country with low cycle helmet use, so the analyses do not reflect any benefit from helmets.)

These data contradict the often stated carricature that cyclists are at special risk. Pedestrians are, indeed, vulnerable road users, but there is no factual justification for the view that cyclists face similar risks. The term 'vulnerable' is often used casually, without being defined or justified by factual information.


Objective risk assessment has been lacking when considering the safety of cyclists, and as a consequence the debate has been unproductive and poorly-informed.

The promotion of helmets for cyclists but not pedestrians or motorists is based upon false assumptions about risk when cycling, since risk assessment reveals that cycling is not a more risky activity than the other modes of transport. Therefore helmets, cannot be more important for cyclists than they are for other road users.

On the other hand, risk assessment does provide evidence of the much greater vulnerability of pedestrians.


Carre, 1992

Carre JR, 1992. La situation de la bicyclette en France. Perspectives Mondiales sur la velo, Velo Quebec, Montreal, 1992 pp 49-54.

Carre, 1995

Carre JR, 1995. La bicyclette: un mode de deplacement meconnu dans ses risques comme dans son usage. Recherche Transports Securite Dec1995;49:19-34.

Krag, 2005

Krag T, . Cycling, safety and health. European Cyclists Federation.

Morgan 1988

Morgan J, 1988. Risk in cycling. TRRL WP/RS/75.

Shafer, 2000

Shafer A, 2000. Regularities in Travel Demand, an International Perspective. Journal of Transportation and Statistics Vol 3 No 3 Dec 2000.

Wardlaw, 2002

Wardlaw MJ, 2002. Assessing the actual risks faced by cyclists. Traffic Engineering + Control Dec 2002 p352-356.