It is generally accepted that the incidence of whiplash associated disorders is increasing in all industrialised countries, despite the almost universal fitment of head restraints in at least the front seats of cars. This is usually attributed to the fact that few people can be observed to follow the standard recommendations as regards head restraint positioning, that is, level with the head vertically and as close to the head as possible horizontally. This study set out to determine whether any other factors, in addition to head restraint adjustment, could be found which would influence the severity of whiplash injury. This was done by linking medical assessment of real-world accident victims with engineering assessment of the accident vehicles. A random sample of road accident victims suffering from whiplash associated disorder was studied. The vehicles they had been travelling in were examined to assess impact severity and, where possible, measurements were made of seat and head restraint adjustment with the subject sitting in the vehicle. All subjects were interviewed to assess the disability resulting from their injuries, and their progress was followed for 12 months. The results were subjected to statistical analysis to try to determine relationships between severity of injury (as measured by resultant disability) and a number of occupant- and vehicle-related factors. A significant proportion of the sample had suffered lumbar strain injury in addition to whiplash, and these were excluded from the present analysis. Frontal impact victims suffered symptoms indistinguishable from those of rear impact victims. The beneficial effects of good head restraint adjustment could not be clearly demonstrated, and some trends, especially in rear impacts, where the benefits of a well-adjusted restraint should have been very clear, indicated that larger distances from head to restraint were associated with lower disability. The paper discusses these counter-intuitive results and their implications.