Objective To compare the risk of road traffic injury calculated using an exposure measure based on people’s mobility, person-hours travelled (person-hours), with the risk obtained using population census, vehicle fleet and vehicle-kilometres travelled.

Methods The rate of road traffic injury on a working day in Catalonia in 2006 was calculated using the number of people injured from the Register of Accidents and Victims of the National Traffic Authority and as denominator: person-hours travelled, from the 2006 Daily Mobility Survey of Catalonia; population census and vehicle fleet, from the National Statistics Institute; and vehicle-kilometres, from the Ministry of Public Works.

Results Compared with person-hours travelled: population census may underestimate the risk in groups with low mobility; vehicle-kilometres may underestimate the risk in regions with high level of non-motorised mobility and high use of public transport; vehicle fleet may underestimate the risk for collective forms of transport such as buses and for motorcyclists who make many trips but of short duration.

Conclusions Measures of exposure involving people's mobility should be used in the estimation of road traffic injury risk, instead of vehicle's mobility, population census or vehicle fleet.