It has been established that seat belt use by car occupants and helmet use by motorcycle riders substantially reduces the risk of serious and fatal injuries following accidents. No study, however, has evaluated the motor vehicle deaths that could be prevented in Greece by general use of these devices, even though this country has the highest mortality from motor vehicle accidents in the European Union. We have estimated the odds ratios (OR) for death rather than injury in a motor vehicle accident by seat belt use among occupants of passenger cars or helmet use among motorcycle riders, using a nationwide database in which persons killed or injured in road traffic accidents in 1985 and 1994 were recorded. The study base included 910 dead and 19,511 injured persons for 1985 and 1203 dead and 22,186 injured persons for 1994. The OR and 95% confidence intervals (CI) for death rather than injury was 0.69 (CI: 0.58 to 0.81, p < 10(-5)) for seat belt users versus non-users and 0.64 (CI: 0.51 to 0.81; p < 10(-3)) for helmet users versus non-users. There was evidence that the protective effect of these passive safety devices increased from 1985 to 1994 probably reflecting technological improvements. The proportion of all deaths that could have been avoided if all car occupants used seat belts was estimated to 27%, whereas 38% of motorcycle deaths could have been avoided if all motorcycle riders used helmets. These proportions translate to about 500 deaths per year, mostly deaths among young men.